Hitachi Command Suite Dynamic Link Manager Software User Guide for Solaris® Document Organization Product Version Getting Help Contents MK-92DLM114-27
© 2011- Hitachi, Ltd. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying and recording, or stored in a database or retrieval system for any purpose without the express written permission of Hitachi, Ltd. Hitachi, Ltd., reserves the right to make changes to this document at any time without notice and assumes no responsibility for its use.
Contents Preface.................................................................................................xiii Intended audience...................................................................................................xiv Product version.......................................................................................................xiv Release notes..........................................................................................................xiv Document revision level.....
Path Status Transition.....................................................................................2-19 The Online Path Status..........................................................................2-19 The Offline Path Status..........................................................................2-20 Status Transitions of a Path....................................................................2-20 Intermittent Error Monitoring (Functionality When Automatic Failback Is Used).........
Notes on Hardware Settings............................................................................3-39 Notes on Installation.......................................................................................3-40 Notes on Related Software..............................................................................3-44 Notes on Command Execution.........................................................................3-45 Notes on the Disk Label.........................................................
Configuring a Boot Disk Environment in Which HDLM Manages the Boot Disk and Mirroring the Environment by Using SVM........................................................3-116 Configuring a Boot Disk Environment in Which HDLM Manages the Boot Disk, from the Local Boot Disk Environment...................................................3-116 Mirroring a Boot Disk Environment in Which HDLM Manages the Boot Disk by Using SVM........................................................................................
Setting Up SDS...................................................................................................3-170 Notes...........................................................................................................3-170 Registering HDLM Devices.............................................................................3-171 To Use a Local Metadevice...................................................................3-171 To Use a Shared Diskset...............................................
Changing the Status of Paths to Online.....................................................4-7 Changing the Status of Paths to Offline(C)................................................4-8 Viewing LU Information.....................................................................................4-9 Displaying Corresponding Information About an HDLM Device, sd or ssd Device, and LDEV...............................................................................................................
5 Troubleshooting....................................................................................5-1 Information collected by using the DLMgetras utility for collecting HDLM error information.............................................................................................................5-2 Checking Error Information in Messages................................................................... 5-2 What To Do for a Path Error..................................................................
To display the format of the set operation...............................................6-18 Parameters.....................................................................................................6-18 To set up the HDLM operating environment............................................6-18 To display the format of the set operation...............................................6-31 view (Displays Information)...................................................................................
The dlmlisthdev Utility for Assisting HDLM Transitions..............................................7-20 Format...........................................................................................................7-20 Parameters.....................................................................................................7-20 The dlmsetboot Utility for Assisting Configuration of an HDLM Boot Disk Environment7-22 Format...............................................................................
Functional Differences Between Version 6.0 or Later and Versions Earlier Than 6.0.....B-2 Precautions on Differences in Functionality Between HDLM 5.6.1 or Earlier and HDLM 5.6.2 or Later.........................................................................................................
Preface This document describes how to use the Hitachi Dynamic Link Manager.
Intended audience This document is intended for storage administrators who use Hitachi Dynamic Link Manager (HDLM) to operate and manage storage systems, and assumes that readers have: • Knowledge of Solaris and its management functionality • Knowledge of Storage system management functionality • Knowledge of Cluster software functionality • Knowledge of Volume management software functionality Product version This document revision applies to HDLM for Solaris version 7.6 or later.
Chapter/Appendix Description Chapter 1, Overview of HDLM on page 1-1 Gives an overview of HDLM, and describes its features. Chapter 2, HDLM Functions on page 2-1 Describes management targets and the system configuration of HDLM, and the basic terms and functions for HDLM. Chapter 3, Creating an HDLM Environment on page 3-1 Describes the procedures for setting up an HDLM environment and the procedure for canceling those settings.
• Hitachi Workgroup Modular Storage Series User's Guide • Thunder9580V Series Disk Array Subsystem User's Guide • Hitachi Universal Storage Platform V Hitachi Universal Storage Platform VM User and Reference Guide • Hitachi Virtual Storage Platform Series User's Guide • Reference Manual / File Conversion Utility & File Access Library Document conventions This document uses the following typographic conventions: Convention Description Bold Indicates text on a window, other than the window title,
Physical capacity unit Value 1 terabyte (TB) 1,000 GB or 1,0004 bytes 1 petabyte (PB) 1,000 TB or 1,0005 bytes 1 exabyte (EB) 1,000 PB or 1,0006 bytes Logical storage capacity values (for example, logical device capacity) are calculated based on the following values: Logical capacity unit Value 1 block 512 bytes 1 KB 1,024 (210) bytes 1 MB 1,024 KB or 1,0242 bytes 1 GB 1,024 MB or 1,0243 bytes 1 TB 1,024 GB or 1,0244 bytes 1 PB 1,024 TB or 1,0245 bytes 1 EB 1,024 PB or 1,0246 bytes A
xviii Preface Hitachi Dynamic Link Manager User Guide for Solaris®
1 Overview of HDLM HDLM is a software package that manages paths between a host and a storage system. HDLM is designed to distribute loads across multiple paths and will switch a given load to another path if there is a failure in the path that is currently being used, thus improving system reliability. This chapter gives an overview of HDLM and describes its features.
What is HDLM? With the widespread use of data warehousing and increasing use of multimedia data, the need for high-speed processing of large volumes of data on networks has rapidly grown. To satisfy this need, networks dedicated to the transfer of data, such as SANs, are now being used to provide access to storage systems. HDLM manages the access paths to these storage systems.
The ability to continue running operations between a host and storage system, even if there is a failure. This is also known as performing a failover. When a host is connected to a storage system via multiple paths, HDLM can automatically switch to another path if there is some sort of failure in the path that is currently being used. This allows operations to continue between a host and a storage system.
1-4 Overview of HDLM Hitachi Dynamic Link Manager User Guide for Solaris®
2 HDLM Functions This chapter describes the various functions that are built into HDLM. Before the function specifications are explained though, this chapter will go into detail about the HDLM management targets, system configuration, and basic terms that are necessary to know to effectively operate HDLM. After that, the rest of the chapter focus on describing all the HDLM functions, including the main ones: load distribution across paths and path switching.
□ Collecting Audit Log Data □ Integrated HDLM management using Global Link Manager □ Cluster Support 2-2 HDLM Functions Hitachi Dynamic Link Manager User Guide for Solaris®
Devices Managed by HDLM Below is a list of devices that can or cannot be managed by HDLM. The devices that can be managed by HDLM are called HDLM management-target devices. HDLM management-target devices: The following devices are from the storage systems listed in Section What is HDLM? on page 1-2: ¢ SCSI devices (sd or ssd devices) ¢ Boot disks# ¢ Swap devices# ¢ Dump devices# #: If you want to use these disks as HDLM management-target devices, assign VTOC labels to them.
paths. The ID that HDLM assigns to each path (physical path) is called a AutoPATH_ID. A path is also sometimes called a managed object. The following figure shows the HDLM system configuration. Figure 2-1 HDLM System Configuration The following table lists and describes the HDLM system components. Table 2-1 HDLM System Components Components 2-4 Description HBA A host bus adapter. This serves as a cable port on the host.
LU Configuration After you have properly installed HDLM, the LU configuration will change as follows: Before the installation of HDLM: The host recognizes that an sd or ssd device is connected to each physical path. Thus, a single LU in the storage system is recognized as the same number of LUs as that of physical paths. After the installation of HDLM: An HDLM device that corresponds one-to-one with the Dev in an LU in the storage system is created above an sd or ssd device.
Figure 2-2 LU Configuration Recognized by the Host After the Installation of HDLM The following table lists and describes the components recognized by the host. Table 2-2 Components Recognized by the Host Components Description HLU An LU that the host recognizes via the HDLM driver. It is called a host LU. No matter how many physical paths exist, one host LU is recognized for one LU in the storage system. HDev A Dev (a slice or partition) in an LU that the host recognizes via the HDLM driver.
Figure 2-3 Configuration of the HDLM Programs The following table lists and describes the functions of these programs.
Program name Functions • Request path health checks and automatic failbacks to be performed • Collect error log data HDLM alert driver Reports the log information collected by the HDLM driver to the HDLM manager. The driver name is dlmadrv. HDLM driver Controls all the HDLM functions, manages paths, and detects errors. The HDLM driver consists of the following: • Core logic component • Filter component Controls the basic functionality of HDLM. Sends and receives I/O data.
Figure 2-4 Position of the HDLM Driver and HDLM Devices Logical Device Files for HDLM Devices When you install HDLM, a logical device file to be used by HDLM will be created for each LU on a per-Dev (slice) basis. Setting this logical device file name in an application, such as volume management software, enables the application to access an LU by using the HDLM function. The logical device files existing before HDLM installation (the logical device files of an sd or ssd) will be deleted.
and for one of the Dev (slices) in that LU, the corresponding logical device file names of the sd or ssd devices are c2t1d1s0 and c3t2d1s0. In this case, when you install HDLM, these logical device files will be deleted. Then, a logical device file that has a different controller number, such as c4t1d1s0, is created for the HDLM device.
Figure 2-5 Flow of I/O Data When the Load Balancing Function Is Not Used When the load balancing function is not used, I/O operations converge onto a single path (A). The load on that one physical path (A) will cause a bottleneck, which might cause problems with system performance.
Figure 2-6 Flow of I/O Data When the Load Balancing Function Is Used When the load balancing function is used, I/O operations are distributed via multiple physical paths (A, B, C, and D). This helps to prevent problems with system performance and helps prevent bottlenecks from occurring. Paths to Which Load Balancing Is Applied This section describes, for each type of storage system, the paths to which the load balancing function is applied.
remaining usable owner paths. When all owner paths cannot be used, load balancing is performed among the non-owner paths. For the example in Figure 2-7 Overview of Load Balancing on page 2-13, suppose that in the owner controller of LU0 is CHA0. When the LU is accessed, the load is balanced between the two physical paths A and B, which are both owner paths. When one of the paths (A) cannot be used, then the LU is accessed from the only other owner physical path (B).
¢ Hitachi AMS2000 series# ¢ Hitachi SMS series# ¢ HUS100 series# HUS VM #: This storage system applies when the dynamic I/O path control function is disabled.
The paths are simply selected in order from among all the connected paths. ¢ ¢ Extended Least I/Os The path that has the least number of I/Os being processed is selected from among all the connected paths. Extended Least Blocks The path that has the least number of I/O blocks being processed is selected from among all the connected paths. The following table lists and describes the features of the load balancing algorithms.
If a path in which an error has occurred recovers from the error, HDLM can then switch back to that path. This is called a failback. Two types of failovers and failbacks are available: • Automatic failovers and failbacks • Manual failovers and failbacks Failovers and failbacks switch which path is being used and also change the statuses of the paths. A path status is either online or offline. An online status means that the path can receive I/Os.
For example, in Figure 2-8 Path Switching on page 2-17, the owner controller of LU is CHA0, and access to the LU is made only via the physical path (A). After the access path is placed offline, the first candidate for the switching destination is the physical path (B) and the second candidate is the physical path (C or D).
Automatic Failbacks When a path recovers from an error, HDLM can automatically place the recovered path back online. This function is called the automatic failback function. In order to use the automatic failback function, HDLM must already be monitoring error recovery on a regular basis. When the Thunder 9500V series, or Hitachi AMS/WMS series is being used, HDLM selects the path to use from online owner paths, and then from online non-owner paths.
However, if there is only one online path for a particular LU, that path cannot be manually switched offline. Also, a path with an error that has not been recovered from yet cannot be switched online. HDLM uses the same algorithms to select the path that will be used next, regardless of whether automatic or manual path switching is used. When the Thunder 9500V series, or Hitachi AMS/WMS series is being used, HDLM selects the switching destination path from owner paths and then from non-owner paths.
If none of the paths accessing a single LU are in the Online status, one of the paths will be changed to the Online(E) status. All the paths that access the same LU will not be in the offline statuses. This ensures access to the LU. The text (E) of Online(E) indicates the error attribute, which indicates that an error occurred in the path. • Online(S)# The paths to the primary volume (P-VOL) in the HAM environment have recovered from an error, but I/O to the P-VOL is suppressed.
Figure 2-9 Path Status Transitions Legend: Online operation: Online operation performed by executing the dlnkmgr command's online operation. Offline operation: Offline operation performed by executing the dlnkmgr command's offline operation. #1 When no Online or Offline(E) paths exist among the paths that access the same LU.
#4 One of the Offline(E) paths is changed to the Online(E) path. #5 When an Offline(E) path exists among the paths that access the same LU. Figure 2-10 Path Status Transitions (P-VOL in HAM environment) Legend: Online operation: Online operation performed by executing the dlnkmgr command's online operation. Offline operation: Offline operation performed by executing the dlnkmgr command's offline operation. #1 Also when an error occurs in all the paths to an S-VOL in the Online(D) status.
When you are using intermittent error monitoring, the path in which the intermittent error occurred is not automatically placed online even when the path recovers from the error. In such a case, place the path online manually. Note If there is a path failure immediately after a path is placed offline by using the dlnkmgr command, Offline(C) might change to Offline(E).
determined to have an intermittent error if an error occurs 3 or more times in 30 minutes. You can set up intermittent error monitoring by executing the dlnkmgr command's set operation. Intermittent error monitoring can be used only when automatic failback has already been enabled. The values that can be specified for intermittent error monitoring depend on the values specified for automatic failbacks. For details on how to specify the settings, see set (Sets Up the Operating Environment) on page 6-17.
Figure 2-11 Action What Will Happen When an Intermittent Error Occurs on a Path When an Intermittent Error Does Not Occur If an error does not occur in the path the specified number of times within the specified interval, the system determines that the path does not have an intermittent error. In such a case, the error monitoring will finish when the specified error-monitoring interval finishes, upon which the number of errors is reset to 0.
As shown in Figure 2-12 What Will Happen When an Intermittent Error Does Not Occur on a Path on page 2-25, normally, the count for the number of times that an error occurs is started after the path is first recovered from an error by using the automatic failback function.
When a User Changes the Intermittent Error Information The following might be reset when a user changes any of the values set for the intermittent error or the path status: the number of errors that have already been counted during error monitoring, the amount of time that has passed since error monitoring has started, and the information about whether an intermittent error has occurred.
User operation Number of errors and time passed since error monitoring started monitoring is being performed Information about paths not subject to automatic failback If a path has been removed from the paths subject to automatic monitoring, that path is no longer monitored. Restarting the HDLM manager Reset#3 Inherited Restarting the host Reset Reset #1 When you disable the intermittent error monitoring function, information about paths not subject to automatic failback will be reset.
connected to non-owner paths, we recommend that you use path health checking to detect errors. This enables the system to use the most up-to-date path-status information when selecting the next path to use. You can configure path health checking by executing the dlnkmgr command's set operation. For details on the set operation, see set (Sets Up the Operating Environment) on page 6-17.
Error Management For troubleshooting purposes, HDLM collects information and stores it into log files. The error information to be collected can be filtered out by error level, and then stored into the log files. The following figure shows the flow of data when error information is collected on a host which is running HDLM. Figure 2-14 Flow of Data When Collecting Error Information Logs might be collected in layers below HDLM, such as for the SCSI driver. For more details, see the Solaris documentation.
Types of Collected Logs HDLM collects information on detected errors and trace information in the integrated trace file, trace file, error logs, log for the dlmcfgmgr utility for managing the HDLM configuration, and syslog. You can use the error information to examine the status of an error and analyze the cause of the error. The following table lists and describes the error information that can be collected in logs.
Log name Description Output destination We recommend that you configure the system so that information at the Information level and higher is output. The syslog file path is specified in the file /etc/ syslog.conf or /etc/ rsyslog.conf. For details, Syslogs can be checked using a text refer to the Solaris documentation. editor. # When you want to configure the system so that HDLM messages are output to syslog, specify user for the facility in the /etc/syslog.conf or /etc/rsyslog.conf file.
recommended that you set the information to be output at the info level and higher. Note that the facility and level (in facility format) when HDLM outputs messages to syslog are all user. The error information in error logs and trace files are collected based on a user-defined collection level. The collection levels are as follows: Collection levels for error logs ¢ Collects no error information. ¢ Collects error information from the Error level and higher.
Table 2-8 Categories of Audit Log Data that Can Be Collected Category StartStop Explanation An event indicating the startup or termination of hardware or software, including: • OS startup and termination • Startup and termination of hardware components (including micro-program) • Startup and termination of software running on storage systems, software running on SVPs (service processors), and Hitachi Command Suite products An abnormal hardware or software event, including: Failure LinkStatus • Ha
Category Explanation • Maintenance AnomalyEvent Viewing or updating audit log settings An event indicating that a maintenance operation has terminated normally or failed, including: • Adding or removing hardware components • Adding or removing software components An event indicating an abnormal state such as exceeding a threshold, including: • Exceeding a network traffic threshold • Exceeding a CPU load threshold • Reporting that the temporary audit log data saved internally is close to its ma
Category Explanation Audit event Severity# 1 Message ID monitoring function was successful. Authentication ConfigurationAccess 2-36 Administrator or end-user authentication Viewing or updating configuration information Startup of the I/O 3 information monitoring function failed. KAPL15113-E The I/O 6 information monitoring function stopped. KAPL15114-I The I/O 4 information monitoring function terminated.
Category Explanation Audit event Severity# 1 Message ID Setup of the 3 operating environment failed. KAPL15106-E An attempt to display program information was successful. 6 KAPL15107-I An attempt to display program information failed. 3 KAPL15108-E An attempt to display HDLM managementtarget information was successful. 6 KAPL15109-I An attempt to display HDLM managementtarget information failed. 3 KAPL15110-E Processing of the dlmcfgmgr -a command was successful.
Category Explanation Audit event Severity# 1 Message ID Processing of the dlmvxexclude [d] command failed. 3 KAPL15027-E The status of a path was successfully changed to Online. 6 KAPL15116-I A path was 6 successfully added. KAPL15117-I Path addition failed.
¢ If the severity specified by the HDLM command's set operation differs from the severity specified by the configuration file /etc/syslog.conf or /etc/rsyslog.conf, the higher severity level is used for outputting audit log data. Destination and Filtering of Audit Log Data Audit log data is output to syslog. Because HDLM messages other than audit log data are also output to syslog, we recommend that you specify the output destination that is used exclusively for audit log data.
¢ All of the above For details on how to specify audit log settings, see Setting Up the HDLM Functions on page 3-134. Audit Log Data Formats The following describes the format of audit log data: Format of audit log data output to syslog: ¢ priority ¢ date-and-time ¢ host-name ¢ program-name ¢ [process-ID] ¢ message-section The following shows the format of message-section and explains its contents.
Item# Explanation Audit event type Event type Audit event result Event result Subject ID for audit event result Depending on the event, an account ID, process ID, or IP address is output.
For large-scale system configurations using many hosts running HDLM, the operational load for managing paths on individual hosts increases with the size of the configuration. By linking HDLM and Global Link Manager, you can centrally manage path information for multiple instances of HDLM and reduce operational load. In addition, you can switch the operational status of paths to perform system-wide load balancing, and centrally manage the system by collecting HDLM failure information in Global Link Manager.
For details on cluster software supported by HDLM, the supported Solaris version, and usable volume management software, see Combinations of Cluster Software and Volume Managers Supported by HDLM on page 3-7. HDLM uses a path of the active host to access an LU. Paths are switched in units of physical paths. Therefore, if an error occurs in a path, all the other paths that run through the same physical path are switched. The switching destination is a physical path of the active host.
2-44 HDLM Functions Hitachi Dynamic Link Manager User Guide for Solaris®
3 Creating an HDLM Environment This chapter describes the procedures for setting up an HDLM environment and the procedure for canceling those settings. Make sure that HDLM is installed and the functions have been set up. Set up volume groups and cluster software to suit your operating environment.
□ Creating File Systems for HDLM (When Volume Management Software Is Not Used) □ Setting Up VxVM □ Setting Up SDS □ Setting Up SVM □ Setting Up VCS □ Removing HDLM 3-2 Creating an HDLM Environment Hitachi Dynamic Link Manager User Guide for Solaris®
HDLM System Requirements Check the following before installing HDLM: For the requirements for using HDLM in an HAM environment, see the release notes of HDLM. Hosts and OSs Supported by HDLM HDLM can be installed on a SPARC series computer which is running an OS listed in the following table.
Table 3-2 JDK required for linking with Global Link Manager OS Solaris 8 or Solaris 9 Solaris 10 Solaris 11 JDK package • JDK 1.4.2_15 or later (32 bit version) • JDK 5.0_11 or later (32 bit version) • JDK 6.0_17 or later (32 bit version) • JDK 1.4.2_15 or later (32 bit version) • JDK 5.0_11 or later (32 bit version) • JDK 6.0_17 or later (32 bit version) • JDK 7.0 (32 bit version) • JDK 6.0_17 or later (32 bit version) • JDK 7.
For details on the micro-programs and settings information for storage systems, which are required to use HDLM, see the HDLM Release Notes and maintenance documentation for the storage system.
Table 3-3 Supported cluster software versions Cluster software Supported versions Local boot disk environment#1 Boot disk environment#2 VCS#3 VCS5.0#4 VCS5.0#4 Sun Cluster#5 Sun Cluster 3.1, Sun Cluster 3.2, Oracle Solaris Cluster 3.3, or Oracle Solaris Cluster 4.0 Sun Cluster 3.1 8/05 (Update 4) Oracle RAC#6 Oracle9i RAC, Oracle RAC 10g, or Oracle RAC 11g -- #1 An environment with a boot disk located on the host.
HDLM-managed boot disks do not support a mirrored boot disk configuration incorporating a volume manager such as SDS or VxVM. For this reason, you cannot register an HDLM-managed boot disk in bootdg when using VxVM.
OS Cluster Sun Cluster 3.1 (8/05)#1 Volume manager None SDS 4.2.1 VCS 5.0#2 #3 None VxVM 5.0 Solaris 9 None None SVM#4 VxVM 5.0#5 Sun Cluster 3.1#1 #5 None SVM#6 Sun Cluster 3.1 (9/04)#1 #5 None SVM#6 Sun Cluster 3.1 (8/05)#1 #5 None SVM#6 VCS 5.0#2 #3 #5 None VxVM 5.0 Solaris 10 None None SVM 1.0#4 #7 #8 VxVM 5.0#5 #8 Sun Cluster 3.1 (8/05) #1 #5 #8 None SVM#6 #9 Sun Cluster 3.2 #1 #10 None#11 SVM#5 #6 #8 #9 VxVM 5.0#5 #8 #12 Sun Cluster 3.
OS Cluster Oracle Solaris Cluster 3.3#10 #13 Volume manager None SVM#5 #6 #8 #9 VxVM 5.1#5 #8 VCS 5.0#3 None#14 VxVM 5.0#5 #8 #15 VCS 5.1#14 None VxVM 5.1#5 #8 Solaris 11 VCS 6.0#16 VxVM 6.0#5 #8 Oracle Solaris Cluster 4.0 None SVM VCS 6.0 VxVM 6.
#7 In a configuration that uses a driver other than the Oracle HBA driver (other than the qlc or emlxs driver), the SVM shared diskset cannot use disks managed by HDLM. #8 Does not support ZFS. #9 When the SVM shared diskset uses disks managed by HDLM in a configuration that uses a driver other than the Oracle HBA driver (other than the qlc or emlxs driver), use Sun Cluster device IDs (logical device files under /dev/did/dsk). The SVM shared diskset cannot use HDLM logical device names.
When Creating an Oracle9i RAC Environment Required programs The following table lists the programs required to create an Oracle9i RAC environment. Table 3-5 Programs required to create an Oracle9i RAC environment (for Solaris 10) Program Remarks OS Solaris 10 -- Cluster Sun Cluster 3.1 8/05 HDLM supports the two-node configuration only. Sun Cluster Support for Oracle Parallel Server/Real Application Clusters 3.
Configurati on Program Remarks Oracle RAC 10g Oracle 10g Database 10.1.0.2.0 -- Cluster Oracle Cluster Ready Services (CRS) 10.1.0.2.0 -- Volume Manager ASM • ASM is bundled with Oracle RAC 10g. • ASM is used as the disk memory area for files and recovery files of the Oracle database. In Oracle RAC 10g, HDLM devices can be used following the same procedures as for disk devices. For details on how to install ASM, refer to the documentation for Oracle RAC 10g.
Configurati on Program Volume Manager None (Specify an HDLM raw device) Configuration OS Solaris 9 4 Oracle RAC 10g Oracle 10g Database 10.1.0.5.0 Remarks ---- Cluster Oracle Clusterware 10.1.0.5.0 -- Volume Manager None (Specify an HDLM raw device) -- Configuration OS Solaris 9 5 Oracle RAC 10g Oracle 10g Database 10.2.0.2.0 --- Cluster Oracle Clusterware 10.2.0.2.0 -- Volume Manager None (Specify an HDLM raw device) -- Configuration OS Solaris 9 6 Oracle RAC 10g Oracle 10g Database 10.
Configurati on Program Remarks the documentation for Oracle RAC 10g. Table 3-7 Programs required to create an Oracle RAC 10g environment (For Solaris 10) Configurati on Configuratio n1 Program Remarks OS Solaris 10 -- Oracle RAC 10g Oracle 10g Database 10.1.0.4.0 -- Cluster Oracle Cluster Ready Services (CRS) 10.1.0.4.0 -- Volume Manager ASM • ASM is bundled with Oracle RAC 10g. • ASM is used as the disk memory area for files and recovery files of the Oracle database.
Configurati on Program Remarks and recovery files of the Oracle database. In Oracle RAC 10g, HDLM devices can be used following the same procedures as for disk devices. For details on how to install ASM, refer to the documentation for Oracle RAC 10g. Configuratio n4 OS Solaris 10 -- Oracle RAC 10g Oracle 10g Database 10.2.0.2.0 -- Cluster Oracle Clusterware 10.2.0.2.0 -- Volume Manager ASM • ASM is bundled with Oracle RAC 10g.
Configurati on Configuratio n7 Configuratio n8 Configuratio n9 Program Remarks OS Solaris 10 -- Oracle RAC 10g Oracle 10g Database 10.2.0.2.0 Cluster Sun Cluster 3.1 8/05 and Oracle Clusterware 10.2.0.2.0 Only two-node configurations are supported. Volume Manager VxVM 4.1 cluster functionality Allocates memory areas, shared among nodes, such as Oracle database files, SPFILE, REDO log files, Oracle Cluster Registry, and voting disks, to the VxVM 4.1 cluster functionality volumes.
Configurati on Configuratio n 11 Program Remarks Oracle RAC 10g Oracle 10g Database 10.2.0.3.0 -- Cluster Sun Cluster 3.2 and Oracle Clusterware 10.2.0.3.0 -- Volume Manager None (Specify an HDLM raw device from the Sun Cluster device ID) -- OS Solaris 10 -- Oracle RAC 10g Oracle 10g Database 10.2.0.3.0 Cluster Sun Cluster 3.2 and Oracle Clusterware 10.2.0.3.0 Only configurations that consist of three or more nodes are supported.
Configurati on Configuratio n 14 Configuratio n 15 Configuratio n 16 Program Remarks Volume Manager VxVM 5.0 cluster functionality# Allocates memory areas, shared among nodes, such as Oracle database files, SPFILE, REDO log files, Oracle Cluster Registry, and voting disks, to the VxVM 5.0 cluster functionality volumes. For details on how to allocate memory areas, refer to the documentation for Oracle RAC 10g. OS Solaris 10 -- Oracle RAC 10g Oracle 10g Database 10.2.0.4.
Configurati on Configuratio n 17 Configuratio n 18 Configuratio n 19 Configuratio n 20 Program Remarks OS Solaris 10 -- Oracle RAC 10g Oracle 10g Database 10.2.0.3.0 Cluster Oracle Solaris Cluster 3.3 and Oracle Clusterware 10.2.0.3.0 Only two-node configurations are supported. Volume Manager VxVM 5.1 cluster functionality# Allocates memory areas, shared among nodes, such as Oracle database files, SPFILE, REDO log files, Oracle Cluster Registry, and voting disks, to the VxVM 5.
Configurati on Program Remarks MP version is the same as that of VCS) are supported. Cluster VCS 5.0# The parallel service group with an I/O fencing function enabled is supported. Volume Manager VxVM 5.0 cluster functionality# Allocates memory areas, shared among nodes, such as Oracle database files, SPFILE, REDO log files, Oracle Cluster Registry, and voting disks, to the VxVM 5.0 cluster functionality volumes.
Target program Patch ID Timing for applying 110386-03 or later 111023-02 or later 111111-03 or later 111308-03 or later 111310-01 or later 112396-02 or later 111721-04 or later 112003-03 or later 112138-01 or later # When using Oracle RAC 10g 10.2.0.1.0, apply 108993-45, not 108993-19.
Note that, depending on the settings of Oracle RAC 10g, Oracle RAC 10g might determine that a node error has occurred before the failover processing performed by HDLM is completed, and then re-configure the cluster. Therefore, when HDLM manages the paths that are connected to an Oracle RAC 10g voting disk, change the following settings according to your version of Oracle RAC 10g: ¢ When using Oracle RAC 10g version 10.1.0.3.0 or later: Change the value of MISSCOUNT according to the storage system type.
Storage system type • Universal Storage Platform V/VM series • Virtual Storage Platform series • HUS VM • Hitachi AMS2000/AMS/ WMS/SMS series • HUS100 series • Thunder 9500V series Formula for obtaining the value of MISSCOUNT number-of-paths-connected-to-the-voting-disk x 30 seconds In addition to the value of MISSCOUNT shown above, also change the value of DISKTIMEOUT. As with MISSCOUNT, the value to be specified in DISKTIMEOUT is determined by the type of storage system.
¢ voting disk ¢ Oracle Cluster Registry ¢ Oracle database file ¢ system table area ¢ users table area ASM disk to be used for ASM disk group creation In an environment where an HDLM raw device is used as the devices listed above, if the HDLM raw device name does not match between the nodes, create an alias device file of the HDLM raw device in each node by using the following procedure and set the created alias device file in Oracle RAC 10g. ¢ a.
# chown root:oinstall /dev/crs_ocr1 # d. Execute the following command for the created alias device file and check that the major number, minor number, owner, group, and access permission mode is properly set: # ls -l /dev/alias-device-file Execution example: # ls -l /dev/crs_ocr1 crw-r----- 1 root oinstall 307, 1608 date/time /dev/crs_ocr1 # When Creating an Oracle RAC 11g Environment Required programs The following table lists programs required to create an Oracle RAC 11g environment.
Configurati on Program Remarks documentation for Oracle RAC 11g. Configuratio OS Solaris 10 n3 Oracle RAC 11g Oracle 11g Database 11.1.0.6.0 Cluster Sun Cluster 3.1 8/05 and Oracle Clusterware 11.1.0.6.0 Volume Manager None (Specify an HDLM raw device) Configuratio OS Solaris 10 n4 Oracle RAC 11g Oracle 11g Database 11.1.0.6.0 Cluster Sun Cluster 3.1 8/05 and Oracle Clusterware 11.1.0.6.0 Volume Manager ASM -Only two-node configurations are supported.
Configurati on Program Remarks • ASM is used as the disk memory area for the Oracle database files and recovery files. For the disk device used by ASM, specify the Sun Cluster device ID. For details on how to use ASM, refer to the documentation for Oracle RAC 11g. Configuratio OS Solaris 10 n7 Oracle RAC 11g Oracle 11g Database 11.1.0.6.0 Cluster Oracle Solaris Cluster 3.3 and Oracle Clusterware 11.1.0.6.
Configurati on Program Configuratio OS Solaris 10 n 10 Oracle RAC 11g Oracle 11g Database 11.1.0.7.0 Remarks --- Cluster Oracle Clusterware 11.1.0.7.0 -- Volume Manager ASM • ASM is bundled with Oracle RAC 11g. • ASM is used as the disk memory area for the Oracle database files and recovery files. In Oracle RAC 11g, HDLM devices can be used as disk devices by following the usual procedures for HDLM devices. For details on how to use ASM, refer to the documentation for Oracle RAC 11g.
Configurati on Program Remarks Oracle database files and recovery files. In Oracle RAC 11g, HDLM devices can be used as disk devices by following the usual procedures for HDLM devices. For details on how to use ASM, refer to the documentation for Oracle RAC 11g. Configuratio OS Solaris 10 n 13 Oracle RAC 11g Oracle 11g Database 11.2.0.1.0 Cluster Sun Cluster 3.2 and Oracle Grid Infrastructure 11.2.0.1.0 Volume Manager ASM -Only two-node configurations are supported.
Configurati on Program Remarks usual procedures for HDLM devices. For details on how to use ASM, refer to the documentation for Oracle RAC 11g. Configuratio OS Solaris 10 n 15 Oracle RAC 11g Oracle 11g Database 11.2.0.2.0 Cluster Oracle Grid Infrastructure 11.2.0.2.0 Volume Manager ASM -Only two-node configurations are supported. • ASM is bundled with Oracle Grid Infrastructure. • ASM is used as the disk memory area for the Oracle database files and recovery files.
Configurati on Program Configuratio OS Solaris 11 n 17 Oracle RAC 11g Oracle 11g Database 11.2.0.3.0 Cluster Oracle Grid Infrastructure 11.2.0.3.0 Volume Manager ASM Remarks -Only two-node configurations are supported. • ASM is bundled with Oracle Grid Infrastructure. • ASM is used as the disk memory area for the Oracle database files and recovery files. In Oracle RAC 11g, HDLM devices can be used as disk devices by following the usual procedures for HDLM devices.
Note • When a host and an Oracle RAC 11g voting disk are connected by multiple paths, HDLM performs failover processing for those paths (in the same way as for normal paths) when an I/O timeout occurs for one of the paths. Note that, depending on the settings of Oracle RAC 11g, Oracle RAC 10g might determine that a node error has occurred before the failover processing performed by HDLM is completed, and then re-configure the cluster.
Table 3-16 Formula for Calculating DISKTIMEOUT Storage system type Number of paths connecte Formula for obtaining the value of d to the DISKTIMEOUT voting disk • Lightning 9900 series 3 or less You do not need to change the value of DISKTIMEOUT.
Execution example: # ls -lL /dev/rdsk/c10t50060E8005271760d5s0 crw-r----- 1 root sys 307, 1608 date/time /dev/rdsk/ c10t50060E8005271760d5s0 # In this example, the major number is 307 and the minor number is 1608. b. Create an alias device file by executing the following command in each node. An alias device file corresponding to one disk slice should have the same name in all the nodes.
The following describes domains supported by HDLM. • If the control domain and the I/O domains that manage the HBAs for HDLM-managed storage are not separated: ¢ ¢ • HDLM can be installed in the control domain. Only domains including I/O domains that manage the HBAs connected to HDLM-managed storage are supported. HDLM cannot be installed in the guest domain. If the control domain and the I/O domains that manage the HBAs for HDLM-managed storage are separated: ¢ HDLM can be installed in I/O domains.
Legend: Y: Can be used N: Cannot be used Memory and Disk Capacity Requirements This subsection explains the memory and disk capacity requirements. Memory Requirements The host requires 11 MB of memory. Disk Capacity Requirements The following table lists the disk capacity requirements for a host. Table 3-18 Disk capacity requirements for a host Directory Disk capacity requirement /etc 3MB /kernel/drv 2MB + x bytes (x = the-number-of-LUs × the-number-ofphysical-paths × 1.
Number of LUs and Paths Supported in HDLM The following table lists the number of LUs and paths supported in HDLM. Table 3-19 Number of LUs and Paths Supported in HDLM Item Number of LUs OS Supported number Solaris 8, Solaris 9 1 to 256 Solaris 10, Solaris 11 1 to 4096# Number of physical paths per LU Solaris 8, Solaris 9, Solaris 10, Solaris 11 1 to 64 Total number of physical paths Solaris 8, Solaris 9 1 to 4096 Solaris 10, Solaris 11 1 to 8192# When Sun Cluster 3.1, Sun Cluster 3.
Figure 3-1 Flow of HDLM Environment Setup HDLM Installation Types This subsection describes the following types of HDLM installation: new installation, upgrade installation, and re-installation. New installation of HDLM: Installing HDLM in a server on which HDLM has not been installed is called a new installation of HDLM. Upgrade installation of HDLM: Installing a newer version than the existing version without removing the existing version is called an upgrade installation of HDLM.
Notes on Creating an HDLM Environment This section gives points to be noted when creating an HDLM environment. For points to be noted during operation of HDLM, see Notes on Using HDLM on page 4-2. Notes on Hardware Settings For details on the settings of the storage system, see storage system manual and maintenance documentation for the storage system. • When using HBAs that can set up a binding between target IDs and storage-system ports, you must set up the binding before installing HDLM.
Notes on Installation • You cannot edit the HDLM driver configuration definition file (/kernel/ drv/dlmfdrv.conf). To change an HDLM-managed device to a nonHDLM-managed device, use the HDLM device unconfiguration function of the HDLM device. For details on the HDLM device unconfiguration function, see Using the HDLM Device Unconfiguration Function When Performing a New Installation of HDLM on page 3-72 or Specifying Whether a Logical Unit Is To Be Managed by HDLM (When Not Using Sun Cluster) on page 4-46.
¢ SVS is an HDLM management-target device. You are upgrading from HDLM 5.6.1 or earlier. When both these conditions are satisfied, the previous generation of the HDLM driver configuration definition file (/kernel/drv/ dlmfdrv.conf.bak) and HDLM driver unconfiguration definition file (/ opt/DynamicLinkManager/config/dlmfdrv.unconf.bak) are also updated. For details about the dlmfdrv.conf.bak and dlmfdrv.unconf.bak files, see Note in section Parameters on page 7-24.
/var/opt/hitachi/HNTRLib2/spool/hntr2n.log (n indicates a file number) Trace files after upgrade installation or re-installation: /var/opt/DynamicLinkManager/log/hdlmtrn.log (n indicates a file number) • If you interrupted a new installation, check whether the /var/opt/ DynamicLinkManager directory exists.
For details about the commands, refer to Solaris 10 or Solaris 11 documentation. The following table shows Solaris zones supported in HDLM. Table 3-20 Support for the Solaris zones Item • Non-global zone Remarks Installing HDLM Supported Not supported -- Path operations Supported Not supported -- Displaying path statuses Supported Not supported -- Accessing HDLM devices Supported Supported The devices must be allocated to the relevant zone.
desired kernel mode. If you switch the kernel mode from 32 bits to 64 bits when only the 32-bit version of HDLM is installed, you must reinstall HDLM after the switch is completed. For details on how to switch the kernel mode, see Switch the Kernel Mode on page 3-48 in Preparations for a New Installation of HDLM on page 3-45. • When installing HDLM on a host where a Device Manager Agent 5.
You can use zoning of Fibre Channel switches or storage-system's WWN security function to separately use HBA ports. You can also use point-to-point connections to physically separate the HBA ports. Notes on Command Execution When HDLM is used, you cannot specify the -r option in the drvconfig, devlinks, disks, or devfsadm command to reconfigure the HDLM devices on another boot disk. To reconfigure HDLM devices, you must boot them from the applicable boot disk.
option specified, the logical device files of the sd or ssd devices to be managed by HDLM will be deleted. Back up all pertaining to the logical device files under the /dev/dsk and /dev/rdsk directories to another directory. The following shows an example of executing the commands: # mkdir /var/tmp/bkfile1 # ls -l /dev/dsk > /var/tmp/bkfile1/devices1 # ls -l /dev/rdsk > /var/tmp/bkfile1/devices2 5. Unregister the device.
Figure 3-3 An example of how to edit the /etc/vfstab file Comment the shaded portions out by placing a hash mark (#) at the beginning of the line, or delete the line. 7. Unregister the device from the volume management software. If the device is registered in VxVM, SDS, or SVM, cancel the registration. Apply Solaris Patches See Hosts and OSs Supported by HDLM on page 3-3, and then apply any necessary patches to the Solaris OS.
When using an ssd driver, you will not need to edit this file, since the connected disks will be automatically recognized. After editing the /kernel/drv/sd.conf file, restart the host by executing either of the following commands with the reconfiguration option (-r) specified: • # reboot -- -r • boot -r (at the ok prompt) After restarting the host, use the format command to confirm that the sd or ssd device that is to be managed by HDLM is displayed.
To boot at the ok prompt, execute the following command: boot boot-device kernel/sparcv9/unix other-boot-options Set Up the /etc/system File HDLM uses a maximum of 8 KB of the stack for the kernel thread. Add the usage of HDLM to this maximum value, specify that total value as the stack size for the kernel thread, and then install HDLM. In Solaris 10 or Solaris 11, default_stksize has priority over lwp_default_stksize.
local0.info /usr/local/audlog Set Up VxVM To use HDLM and VxVM on the same host, set up VxVM as follows: 1. Make sure that the logical device file under the /dev/dsk and /dev/rdsk directories is backed up. If the information is not backed up, perform step 4 in Performing Operations on Devices to Be Managed by HDLM on page 3-45 in Preparations for a New Installation of HDLM on page 3-45. 2. If a created disk group already contains a device that is to be managed by HDLM, deport the disk group.
• • Release the disk device groups that consist of the sd or ssd devices to be managed by HDLM. For details on how to unregister an SDS or SVM diskset, see Remove SDS Settings on page 3-183 or Remove SVM Settings on page 3-184 in Preparations for HDLM Removal on page 3-179. Unregistration of a diskset deletes the data in the diskset. Therefore, back up a diskset before unregistering it. To remove a VxVM disk group from Sun Cluster's disk device group, perform the following procedures: a.
The following shows an example of executing the command: host2# /usr/cluster/bin/scconf -c -q node=host1,maintstate g. Confirm that the cluster node is in the maintenance mode. Make sure that the node that is in the maintenance node has the following status: - The node is in the Offline status. - Both the current Quorum votes present and Quorum votes possible are 0. The following shows an example of executing the command: host2# /usr/cluster/bin/scstat -q h.
can access the publisher normally, the following procedure is not required. 1. Execute the pkg publisher command. # pkg publisher PUBLISHER TYPE STATUS URI solaris origin online http://pkg.oracle.com/solaris/release/ 2. Disable the publisher of the package currently in use. # pkg set-publisher --disable solaris 3. Install HDLM. 4. Enable the publisher of the package again. # pkg set-publisher --enable solaris Verifying signature-policy HDLM does not support signatures.
solaris shared 1 non-globalzone solaris excl 2. running /zones/non-globalzone If the status is running, shut down the non-global zone and change the status to installed. # zoneadm -z non-globalzone shutdown # zoneadm list -cv ID NAME STATUS PATH BRAND IP 0 global running / solaris shared 1 non-globalzone installed /zones/non-globalzone solaris excl 3. Detach the non-global zone. # zoneadm -z non-globalzone detach 4. Install HDLM. 5. Attach and activate the non-global zone.
For details on unattended installation, see the procedures in Performing a New Installation of HDLM (When Sun Cluster Is Not Being Used) on page 3-55, Performing a New Installation of HDLM (When Sun Cluster Is Being Used) on page 3-62, and Performing an Upgrade Installation or Reinstallation of HDLM on page 3-73. Performing a New Installation of HDLM (When Sun Cluster Is Not Being Used) To perform a new installation of HDLM: 1.
In this example, the shaded portion shows information for each path of the device you are removing from HDLM management. Type in the appropriate device number, and then type quit to end the format command. 5. Insert the DVD-ROM. After a few seconds, the DVD-ROM will be automatically mounted, and the mount point of the DVD-ROM will be displayed in a list of files.
# DVD-ROM-mount-point/installux.sh -f /any-directory/ installation-information-settings-file When executing the installhdlm utility # DVD-ROM-mount-point/HDLM_Solaris/installhdlm -f/anydirectory/installation-information-settings-file If execution of the utility is completed without problems, the KAPL09211-I message is output. If the restart key is set to y (to restart the computer) in the installation-information settings file, make sure that the KAPL09211-I message has been output to the installhdlm.
# rm -rf /var/opt/DynamicLinkManager 8. Execute the following command to check whether HDLM has been installed successfully: For Solaris 8, Solaris 9, or Solaris 10 # pkginfo -l PKGINST: NAME: CATEGORY: ARCH: VERSION: BASEDIR: VENDOR: ... # DLManager DLManager Dynamic Link Manager system sparc xx.x.x.xxxx / If xx.x.x.xxxx is displayed for VERSION, then the correct version of HDLM is installed. The installed version of HDLM is shown for xx.x.x.xxxx.
Type y in response to the KAPL10242-I message displayed when you execute the dlmsetconf utility. Information pertaining to LUs to which the physical path accesses will be defined in the HDLM driver configuration definition file (/kernel/drv/ dlmfdrv.conf). Do not edit this file. 11. Specify the devices that are to be excluded from management by HDLM, if there are any. This step is unnecessary if you want HDLM to manage all applicable devices.
# /opt/DynamicLinkManager/bin/dlnkmgr view -sys HDLM Version : x.x.x-xx Service Pack Version : Load Balance : on(extended lio) Support Cluster : Elog Level : 3 Elog File Size (KB) : 9900 Number Of Elog Files : 2 Trace Level : 0 Trace File Size(KB) : 1000 Number Of Trace Files : 4 Path Health Checking : on(30) Auto Failback : on(1) Intermittent Error Monitor : off Dynamic I/O Path Control : off(10) HDLM Manager Ver WakeupTime Alive x.x.
the HDLM device) and iLU (the internal LU number). Then, compare the correspondence of the logical device file names and internal LUs that took place before HDLM was installed, and then re-register the correct HDLM devices in those programs that will use these devices. 19.
21. Based on the results of step 20, compare the logical device file names from before and after the HDLM installation, and re-register the correct HDLM devices in programs that use HDLM devices. Once installation is complete, check the path information according to the procedures listed in Checking the Path Configuration on page 3-133 to check whether HDLM has been installed correctly or not.
In this example, the shaded portion shows information for each path of the device you are removing from HDLM management. Type in the appropriate device number, and then type quit to end the format command. 5. If the settings of a Quorum device or disk device group are specified for the device to be managed by HDLM, remove those settings.
If no key is displayed, perform the steps beginning with step 11. 7. Stop all nodes that comprise the cluster. On any one node, execute the following command: # /usr/cluster/bin/scshutdown -g0 -y 8. Start all nodes comprising the cluster in non-cluster mode. Execute boot -x at the ok prompt. 9. Cancel the key registration for the device to be managed by HDLM.
cdrom/cdrom0 12. Perform the installation. ¢ If you are not performing an unattended installation of HDLM, execute either of the following installation commands: When executing the installux.sh script # DVD-ROM-mount-point/installux.sh When executing the installhdlm command # DVD-ROM-mount-point/HDLM_Solaris/installhdlm ¢ If you are performing an unattended installation of HDLM, set the installation-information settings file with the -f option specified as shown below. When executing the installux.
¢ When a correct license key is entered, the installation program displays a message that indicates the license was installed. Take care that if you enter an incorrect license key three times, installation stops. If you interrupt an installation, check whether the /var/opt/ DynamicLinkManager directory exists. If the /var/opt/ DynamicLinkManager directory exists, execute the following command to delete it: ¢ # rm -rf /var/opt/DynamicLinkManager 14.
HDLM on page 3-45 in Preparations for a New Installation of HDLM on page 3-45. 16. Execute the following dlmsetconf utility to create the HDLM driver configuration definition file (kernel/drv/dlmfdrv.conf) and the HDLM driver unconfiguration definition file (/opt/DynamicLinkManager/config/ dlmfdrv.unconf). # /opt/DynamicLinkManager/bin/dlmsetconf Type y in response to the KAPL10242-I message displayed when you execute the dlmsetconf utility.
19. Stop all nodes that comprise the cluster. If one or more keys were displayed at step 6 and the node was started in non-cluster mode at step 8, execute the following command on all nodes: # shutdown -g0 -y -i0 If the nodes were started in cluster mode, execute the following command on any one node: # /usr/cluster/bin/scshutdown -g0 -y 20. With the reconfiguration option specified, start all nodes comprising the cluster in cluster mode, at the same time. Execute boot -r at the ok prompt.
= view, completion time = yyyy/mm/dd hh:mm:ss # Nothing is displayed for Support Cluster even if you are using cluster software. However, the cluster-support functions are, in fact, operating without problem. 22. By checking the execution results of the view operation, verify the version of the installed HDLM. If HDLM Version is x.x.x-xx, the installed version of HDLM is correct. The installed version of HDLM is shown for x.x.x-xx. 23.
Table 3-22 Information displayed by the dlmlisthdev utility for assisting HDLM transitions Item Description OldHDevName Indicates the HDLM 04-00 logical device file name (HDevName). If a hyphen (-) is displayed, the corresponding HDev did not exist before the new installation of HDLM. NewHDevName Indicates the logical device file name (HDevName) of the newly installed HDLM. If a hyphen (-) is displayed, the corresponding HDev does not exist after the new installation of HDLM.
¢ In a two-node configuration, execute the following command to release the cluster installation mode: # /usr/cluster/bin/scconf -c -q reset 32. On all nodes in the cluster, add /opt/DynamicLinkManager/bin to the PATH environment variable in the environment settings files (such as .profile) for root permissions.
host1# mount -g /dev/global/dsk/d5s0 /global/mntpt Using the HDLM Device Unconfiguration Function When Performing a New Installation of HDLM You can use the HDLM device unconfiguration function to remove Hitachi storage system devices from HDLM management on a per-LU basis. This section explains how to use the HDLM device unconfiguration function. 1. To remove an LU from HDLM management, use a text editor to edit the HDLM driver unconfiguration definition file (/opt/DynamicLinkManager/ config/dlmfdrv.
# /opt/DynamicLinkManager/bin/dlmsetconf -u 3. 4. Restart the host with the reconfiguration option specified. Execute one of the following commands: ¢ # reboot -- -r ¢ boot -r (at the ok prompt) Execute the following command to verify that the settings were applied correctly: # /opt/DynamicLinkManager/bin/dlnkmgr view -drv Confirm that the LDEV that was removed from HDLM management is not shown.
immediately after the upgrade installation or re-installation has finished. ¢ ¢ 1. Do not interrupt the upgrade installation or re-installation. When you install HDLM while the 32-bit kernel mode is being used, only the 32-bit version of HDLM will be installed. When you install HDLM while the 64-bit kernel mode is being used, both the 32-bit and 64-bit versions of HDLM will be installed.
cdrom/cdrom0 4. Perform the installation. ¢ If you are not performing an unattended installation of HDLM, execute either of the following installation commands: When executing the installux.sh script # DVD-ROM-mount-point/installux.sh When executing the installhdlm utility # DVD-ROM-mount-point/HDLM_Solaris/installhdlm • If you are performing an unattended installation of HDLM, set the installation-information settings file with the -f option specified as shown below. When executing the installux.
¢ ¢ 6. When a correct license key is entered, the installation program displays a message that indicates the license was installed. Take care that if you enter an incorrect license key three times, installation stops. Execute the following command to check that the package is installed, by the version number in the output results: The following shows an example of executing the command: For Solaris 8, Solaris 9, or Solaris 10 # pkginfo -l PKGINST: NAME: CATEGORY: ARCH: VERSION: BASEDIR: VENDOR: ...
# /opt/DynamicLinkManager/bin/dlnkmgr view -sys HDLM Version : x.x.x-xx Service Pack Version : Load Balance : on(rr) Support Cluster : Elog Level : 3 Elog File Size (KB) : 9900 Number Of Elog Files : 2 Trace Level : 0 Trace File Size(KB) : 1000 Number Of Trace Files : 4 Path Health Checking : on(30) Auto Failback : on(1) Intermittent Error Monitor : off Dynamic I/O Path Control : off(10) HDLM Manager Ver WakeupTime Alive x.x.x-xx yyyy/mm/dd hh:mm:ss HDLM Alert Driver Ver WakeupTime ElogMem Size Alive x.x.
Installing HDLM in an LDoms Environment This section explains how to install HDLM in a system configuration in which HDLM-managed sd or ssd devices are assigned as virtual disks for the guest domain. The following figure shows a configuration example before installing HDLM. Figure 3-6 Configuration Example (Before Installation) # A virtual disk backend is a location where virtual disk data is stored. A disk, disk slice, file, or volume (such as ZFS, SVM, and VxVM) can be used for the backend.
Domain Name ldg1 Domain Type An I/O domain (including service domains) Boot Disk vol1@primary-vds0 Exporting Virtual Disk Backends vol1: /dev/dsk/ c2t50060E8010027A82d0s2 (For a local boot disk of the guest domain) vol2: /dev/dsk/ c2t50060E8010027A82d1s2 (For a data disk of the guest domain) ldg2 A guest domain vol1@ldg1-vds0 -- The following figure shows a configuration example for an LDoms environment before installing HDLM.
To install HDLM to a host in which LDoms have been set: 1. Log in to the control domain. 2. Execute the ldm stop command to stop the guest domain. For example, to stop a guest domain ldg2, execute the command as follows: # ldm stop ldg2 3. Execute the ldm unbind command to unbind the resource of the guest domain. For example, to unbind a resource of the guest domain ldg2, execute the command as follows: # ldm unbind ldg2 4.
For details on the HDLM installation procedure, see Performing a New Installation of HDLM (When Sun Cluster Is Not Being Used) on page 3-55 or Performing an Upgrade Installation or Re-installation of HDLM on page 3-73 After installing HDLM, restore the volume group and file system that will be used as the virtual disk backend. 8. On the control domain, execute the ldm add-vdsdev command to export the virtual disk backends from the service domain on the I/O domain to which HDLM will be installed.
Figure 3-7 Configuration Example (After Installation) In this configuration example, /dev/dsk/c4t50060E8010027A82d0s2 and /dev/dsk/c4t50060E8010027A82d1s2 are HDLM devices. Configuring a Boot Disk Environment This section explains how to configure a boot disk environment in which HDLM manages the boot disk. An environment with a boot disk located on the host is called a local boot disk environment.
• To use an existing local boot disk environment without HDLM installed as the pre-migration environment: Perform the procedure described in Migration by Installing HDLM in the Existing Local Boot Disk Environment on page 3-85 in Procedure for Configuring a Boot Disk Environment on page 3-83, and then execute the procedure described in Setting Up the Post-Migration Environment on page 3-91 in Procedure for Configuring a Boot Disk Environment on page 3-83.
Figure 3-8 Flow of setting up the boot disk environment when using an existing HDLM environment as the pre-migration environment 1. Allocate the LUs to be used by the boot disk in the post-migration environment. 2. Get Solaris to recognize the LUs to be used by the post-migration boot disk. 3. Set the LU slices to be used by the post-migration boot disk. 4. Execute the dlmsetconf utility as follows to create the HDLM driver configuration definition file (/kernel/drv/dlmfdrv.
This step and the next step set the LU to be used for the post-migration boot disk as an HDLM-managed LU. 5. Restart the host with the reconfiguration option specified to make the settings valid and the HDLM device recognizable. Execute one of the following commands: ¢ # reboot -- -r ¢ boot -r (at the ok prompt) After you finish these steps, perform the procedure described in Setting Up the Post-Migration Environment on page 3-91.
Figure 3-9 Flow of setting up the boot disk environment when installing HDLM in the existing local boot disk environment 1. Allocate the LUs to be used by the boot disk in the post-migration environment. Configure the LU or LUs to be used by the boot disk as follows: ¢ ¢ 2. 3-86 To the slices of a single LU, allocate the Solaris file system, swap device, dump device, Sun Cluster global device file system, and replica of the SDS or SVM state database.
3. Set the LU slices to be used by the post-migration boot disk. 4. Complete the preparations for installing HDLM in the pre-migration environment and then install HDLM, referring to the description in Installing HDLM on page 3-45. Perform the necessary steps.
Figure 3-10 Flow of setting up the boot disk environment when using an existing boot disk environment without HDLM installed as the premigration environment 1. Allocate the LUs to be used by the boot disk in the post-migration environment. Configure the LU or LUs to be used by the boot disk as follows: ¢ ¢ 3-88 To the slices of a single LU, allocate the Solaris file system, swap device, dump device, Sun Cluster global device file system, and replica of the SDS or SVM state database.
2. Get Solaris to recognize the LUs to be used by the post-migration boot disk. 3. Set the LU slices to be used by the post-migration boot disk. 4. Complete the preparations for installing HDLM in the pre-migration environment and then install HDLM, referring to the description in Installing HDLM on page 3-45. Perform the necessary steps.
Figure 3-11 Flow of setting up the boot disk environment when building a new pre-migration environment 3-90 1. Install Solaris in the pre-migration environment. 2. Install Sun Cluster if required. For details, see your Sun Cluster documentation. 3. Install VxVM if required. For details, see your VxVM documentation. 4. Allocate the LUs to be used by the boot disk in the post-migration environment.
¢ ¢ To the slices of a single LU, allocate the Solaris file system, swap devices, dump devices, Sun Cluster global device file system, and replica of the SDS or SVM state database. To the slices of multiple LUs, allocate the Solaris file system, swap devices, dump devices, Sun Cluster global device file system, and replica of the SDS or SVM state database. In this case, make sure that the physical path of each LU is the same. 5.
# mount /dev/dsk/c8t50060E80004361D1d1s0 /mnt0 # mount /dev/dsk/c8t50060E80004361D1d1s6 /mnt1 # mount /dev/dsk/c8t50060E80004361D1d1s3 /mnt2 4. Copy the Solaris file system from the pre-migration environment to the post-migration environment. If using Sun Cluster, also copy the Sun Cluster global device file system to the post-migration environment.
/dev/dsk/c8t50060E80004361D1d1s0 /dev/rdsk/ c8t50060E80004361D1d1s0 / ufs 1 no /dev/dsk/c8t50060E80004361D1d1s6 /dev/rdsk/ c8t50060E80004361D1d1s6 /usr ufs 1 no /dev/did/dsk/d30s3 /dev/did/rdsk/d30s3 /global/.devices/ node@1 ufs 2 no global # Execute the scdidadm -l command to find out the device ID to specify as the mount device of the Sun Cluster global device file system. In the following example, d24 is the pre-migration boot disk, and d30 is the post-migration boot disk. # scdidadm -l ...
Configuration Is Changed on page 4-17 in Precautions Regarding Changes to the Configuration of an HDLM Operating Environment on page 4-17. To obtain the boot device name specified by OBP, edit the character string of the path determined by the above procedure as follows: - Replace the driver name (ssd or sd) of an ssd or sd driver with disk. - Add the slice identifier of the / (root) file system (:a value from :a to :h) to the end.
¢ ¢ ¢ Perform steps 13 to 17 only if SDS or SVM is being used and you are migrating a replica of the SDS or SVM state database to the postmigration environment. In this case, keep the pre-migration environment recognized by Solaris. Skip to step 18 if SDS or SVM is being used but you are not migrating a replica of the SDS or SVM state database to the post-migration environment. Skip to step 18 if SDS or SVM is not being used. 13. Check the state database settings. An example is shown below. 14.
Perform a backup so that you can take action in the following situations: ¢ ¢ Recover when an error occurs in the post-migration environment. Boot from the pre-migration environment, and repair the postmigration environment. The post-migration backup can be used to restore the post-migration environment. Migrate from a boot disk environment to the local boot disk environment, and return to the state before setting up the boot disk environment.
For Solaris 10 Figure 3-12 Creating a ZFS boot disk environment 1. Copy the local boot disk environment from the ZFS file system to the LUs (SCSI device) in the storage system. 2. Replace the LUs on the SCSI device in the storage system with those on the HDLM device. You can also use the above procedure to create a boot disk when the premigration boot disk is in the storage system rather than on the local disk, or when the file system is a UFS file system.
For Solaris 11 Figure 3-13 Creating a ZFS boot disk environment 1. Migrate the local boot disk environment in ZFS to an LU (HDLM device) in the storage system. 2. Configure the ZFS boot disk environment after the migration. Creating a ZFS Boot Disk Environment (for Solaris 10) The following describes the procedure for creating a ZFS boot disk environment for Solaris 10..
• Two LUs have the same capacity as the local boot disk need to be created in the storage system. You also need to configure at least one data disk on the host. • If HDLM has already been installed, remove either of the two LUs created in the storage system from HDLM management. Set the other LU as an HDLM-managed device. In the procedure below, the LU removed from HDLM management is referred to as LU#0, and the LU set as the HDLM-managed device is referred to as LU#1.
# lustatus Boot Environment Is Active Active Can Copy Name Complete Now On Reboot Delete Status --------------------------------------------------------------s10s_u9wos_14a yes yes yes no zfsBE yes no no yes - 6. Execute the following commands to allow the system to boot in the ZFS boot disk environment.
Replacing the ZFS boot disk environment on the SCSI device with the ZFS boot disk environment on the HDLM device Set up a device for the ZFS boot disk environment created in Boot Disk Environment that uses a ZFS File System on page 3-96 on the HDLMmanaged device. If HDLM has not been installed on the device for the ZFS boot disk environment, install HDLM. After installing HDLM, remove the device for the local boot disk environment from HDLM management, and then perform the procedure below.
4. Execute the installboot command to set the LU#1 to be used for a ZFS boot disk environment. Execute the following command to allow the system to boot in a new boot disk environment. # installboot -F zfs /usr/platform/`uname -i`/lib/fs/zfs/ bootblk HDLM-device-(raw-device)-after-replacement Note that the SCSI device specified for the installboot command must be a raw device. In the following example, the HDLM physical device is /devices/pseudo/ dlmndrv@1/dlmfdrv@w50060e8010253330,1:a.
3. Execute the zpool status command and make sure that the root pool has been created correctly. # zpool status hdlmpool 4. Execute the installboot command to copy the boot block so that you can boot the OS from the ZFS boot disk environment. # installboot -F zfs /usr/platform/`uname -i`/lib/fs/zfs/ bootblk /dev/rdsk/c6t50060E8010027A80d7s0 5. Execute the following commands to create an empty /hdlmpool/boot/ menu.lst. # mkdir /hdlmpool/boot # touch /hdlmpool/boot/menu.lst 6.
# eeprom boot-device=/pci@7c0/pci@0/pci@1/pci@0,2/ SUNW,qlc@2,1/fp@0,0/disk@w50060e8010027a80,7:a For how to obtain a device name to be set for OBP, see the documentation of the HBA you are using. 10. Execute the init command and restart the host. # init 6 Do not use the reboot command to restart the host. 11. Execute the beadm list command to make sure the OS is booted from the boot device you created named hdlmboot.
Assume that the root pool name of the local boot disk environment before the migration is rpool and the root pool name of the ZFS boot disk environment after the migration is hdlmpool. Migrating a swap device and a dump device to a ZFS boot disk environment Migrate a swap device and a dump device to a ZFS boot disk environment. 1. Execute the zfs create command to create a swap device for a file system in the root pool (hdlmpool) after the migration.
Migrating data under the rpool/export directory to a ZFS boot disk environment Migrate data under the rpool/export directory to a ZFS boot disk environment. Before the migration, back up data under the rpool/export directory located in the local boot disk environment before the migration. Otherwise, you will not be able to restarte the system in the local boot disk environment before the migration. 1.
------- ---------- ----- ------ ------hdlmboot NR / 2.18G static 2011-12-06 11:37 solaris R 2.94G static 2011-11-29 14:42 8. Execute the df command to make sure the file system is mounted.
Set the disk slices for an LU as follows: ¢ Slice 0: All disk areas are allocated. Slice 2: All disk areas are allocated. Other slices are not used. ¢ 2. Execute the zpool create command to create a root pool on a built-in disk. # zpool create root-pool-name slice-0-of-internal-disk In the following example, the root pool name is rpool3 and the name of the device on the built-in disk is c1t0d0s0. # zpool create rpool3 c1t0d0s0 3.
do not know the copy procedure, contact the company with which you have a contract for Oracle Support Services. Migrating from a ZFS Boot Disk Environment to the Local Boot Disk Environment (for Solaris 11) This section explains how to migrate from a ZFS boot disk environment to the local boot disk environment.
4. Execute the command below to switch the OS so that the system is started from the OS in the local boot disk environment. # beadm activate solaris # beadm list BE Active Mountpoint ------- ---------hdlmboot R solaris NR / Space Policy Created ---------- ------2.06G static 2011-12-06 11:37 2.93G static 2011-11-29 14:42 Make sure NR is shown in the Active column solaris which you switched to. 5. Execute the eeprom command to configure a device in the local boot disk environment as the boot device.
Figure 3-14 Migration flow when coping the file system from a boot disk environment to the local boot disk environment 1. Boot the system from the pre-migration environment. Perform steps 1 to 16 in the pre-migration environment. 2. Allocate the LUs to be used by the boot disk in the post-migration environment. 3. Get Solaris to recognize the LUs to be used by the post-migration boot disk. 4. Set the LU slices to be used by the post-migration boot disk. ¢ If using Sun Cluster, go to the next step.
5. Assign device IDs. The following shows an example of executing the command: # /usr/cluster/bin/scdidadm -r 6. Notify the server of the new device ID information. The following shows an example of executing the command: # /usr/cluster/bin/scdidadm -ui 7. Update the Sun Cluster global device namespace. The following shows an example of executing the command: # /usr/cluster/bin/scgdevs 8.
rf -) 13. In the post-migration /etc/system file, delete the rootdev parameter that specifies the HDLM physical device file name. 14. Edit the post-migration boot disk's /etc/vfstab file, changing the Solaris file system to the HDLM logical device file name of the post-migration boot disk. If using Sun Cluster, change the mount device# of the Sun Cluster global device file system in the same way. An example is shown below.
Before editing DUMPADM_DEVICE=/dev/dsk/c8t50060E80004361D1d1s1 After editing DUMPADM_DEVICE=/dev/dsk/c0t0d0s1 17. Boot the system from the post-migration environment. ¢ ¢ ¢ Skip to step 18 if SDS or SVM is being used and you are migrating a replica of the SDS or SVM state database to the post-migration environment. In this case, keep the pre-migration environment recognized by Solaris. Skip to step 21 if SDS or SVM is being used but you are not migrating a replica of the SDS or SVM state database.
¢ ¢ VxVM Sun StorageTek Traffic Manager software (or, in Solaris 10, the Solaris multipathing software MPxIO) Configuring a Mirrored Boot Disk Environment Incorporating SVM A mirrored boot disk incorporating SVM can be managed by HDLM. The following configuration is supported: • OS: Solaris 10 • RAID level: Mirroring (no more than three mirrors) • Cluster: None Precautions • When you use the dlmsetconf utility, do not specify the -r option.
• a. Remove HDLM after migration to a local boot disk environment by performing the procedures in Removing HDLM from the Boot Disk Environment on page 3-187 in Removing HDLM on page 3-185. b. Upgrade Solaris. c. Configure a boot disk environment again by performing the procedure in Configuring a Boot Disk Environment in Which HDLM Manages the Boot Disk and Mirroring the Environment by Using SVM on page 3-116.
Figure 3-15 Configuring a boot disk environment in which HDLM manages the boot disk, from the local boot disk environment Mirroring a Boot Disk Environment in Which HDLM Manages the Boot Disk by Using SVM Mirror a configured boot disk by using SVM. The following figure provides an overview of mirroring a boot disk environment.
• Slice 0: / (root) file system • Slice 1: Swap device and dump device • Slice 6: /usr file system • Slice 7: SVM state database 1. Prepare another LU with the same capacity as that of the created boot disk, and make the configuration and capacity of slices the same. 2. Create the SVM state database. # matadb -a -f c8t50060E08000436D1d4s7 # metadb -a c8t50060E08000436D1d5s7 3. Decide the metadevice names. In this procedure, specify the settings as follows: ¢ 4.
7. Execute the metaroot command. # metaroot d10 8. Reboot the host. 9. When the host starts, check the swap device and dump device. Then execute the mount command to confirm that / (root) and /usr have been correctly configured.
Stripe 0: Device Dbase State Reloc Hot Spare c8t50060E08000436D1d4s0 0 d12: Submirror of d10 State: Okay Size: 25190400 blocks (12 GB) Stripe 0: Device Dbase State Reloc Hot Spare c8t50060E08000436D1d5s0 0 ... Start Block /dev/dsk/ No Okay Yes Start Block /dev/dsk/ No Okay Yes 12. Execute the dlnkmgr command's view -path operation to confirm that an I/O is issued to the LU set as the boot disk.
Installing HDLM and then Configuring the Environment 1. Install HDLM by referring to Installing HDLM on page 3-45. 2. Execute the dlmsetconf utility for creating the HDLM driver configuration definition file to configure the HDLM device. 3. Remove the LUs of the mirrored boot disks from HDLM management. For details, see Using the HDLM Device Unconfiguration Function When Performing a New Installation of HDLM on page 3-72. 4. Reboot the host with the reconfiguration option specified.
• Slice 7: SVM state database 1. Decide the metadevice names. In this section, it is assumed that the metadevice names of the existing boot disks and the boot disks after the migration are as shown in following table. Table 3-24 Metadevice names Boot disk Current File system /(root) Mirror d10 Sub mirror d11 d12 swap d20 d21 d22 /usr d30 d31 d32 After migration /(root) d100 d101 d102 swap d200 d201 d202 /usr d300 d301 d302 2.
# newfs /dev/md/rdsk/d300 5. Mount the Solaris basic file system of the post-migration environment. For example: # mount /dev/md/dsk/d100 /mnt0 # mount /dev/md/dsk/d300 /mnt1 6. Copy the existing Solaris basic file system to the post-migration environment. For example: # ufsdump 0f - / | (cd /mnt0; ufsrestore rf -) # ufsdump 0f - /usr | (cd /mnt1; ufsrestore rf -) 7.
To determine this path, perform the procedure in When the Path Configuration Is Changed on page 4-17 in Precautions Regarding Changes to the Configuration of an HDLM Operating Environment on page 4-17. To obtain the boot device name specified by OBP, edit the character string of the determined path as follows: ¢ ¢ Replace the driver name (ssd) of an ssd driver with disk. Add the slice identifier of the / (root) file system (:a value from :a to :h) to the end.
# swap -l swapfile /dev/md/dsk/d200 # dumpadm Dump content: Dump device: Savecore directory: Savecore enabled: dev swaplo blocks free 85,200 16 8409584 8409584 kernel pages /dev/md/dsk/d200 (swap) /var/crash/host yes # mount / on /dev/md/dsk/d100 read/write/setuid/devices/intr/largefiles/ logging/xattr/onerror=panic/dev=1540064 on Mon Mar 9 11:00:25 2009 /usr on /dev/md/dsk/d300 read/write/setuid/devices/intr/ largefiles/logging/xattr/onerror=panic/dev=154012c on Mon Mar 9 11:00:25 2009 13.
Size: 25190400 blocks (12 GB) d101: Submirror of d100 State: Okay Size: 25190400 blocks (12 GB) Stripe 0: Device Dbase State Reloc Hot Spare /dev/dsk/c8t50060E08000436D1d6s0 No Okay Yes d102: Submirror of d100 State: Okay Size: 25190400 blocks (12 GB) Stripe 0: Device Dbase State Reloc Hot Spare /dev/dsk/c8t50060E08000436D1d7s0 No Okay Yes ... Start Block 0 Start Block 0 17. Execute the dlnkmgr command's view -path operation to confirm that I/ O is issued to the LU set as the boot disk.
Backups of the environments are required for the following situations: ¢ ¢ If an error occurs in the post-migration environment: Reboot from the pre-migration environment, and then use it to restore the post-migration environment. Alternatively, the backup of the post-migration environment can also be restored. If you want to migrate from a boot disk environment to a local boot disk environment, you can use the backup to return to the state before creating the boot disk environment.
Figure 3-19 Current boot disks and boot disks after the migration The LU device names of the current boot disks are c8t50060E08000436D1d4 and c8t50060E08000436D1d5, and the LU device names of the boot disk after the migration are c6t50060E08000436D1d6 and c6t50060E08000436D1d7, as shown in the above figure.
Boot disk File system /usr Mirror d300 Sub mirror d301 d302 After migration /(root) d10 d11 d12 swap d20 d21 d22 /usr d30 d31 d32 2. Create metadevices on the sd/ssd device, and then mirror them. (Single mirror) # metainit -f d11 1 1 c6t50060E08000436D1d6s0 # metainit d10 -m d11 # metainit -f d21 1 1 c6t50060E08000436D1d6s1 # metainit d20 -m d21 # metainit -f d31 1 1 c6t50060E08000436D1d6s6 # metainit d30 -m d31 3.
# ufsdump 0f - /usr | (cd /mnt1; ufsrestore rf -) 7. Edit the file /etc/vfstab (/mnt0/etc/vfstab) of the boot disk after the migration to change the definition of the swap after the migration to a metadevice (/dev/md/dsk/d20) and change the definition of /usr after the migration to a metadevice (/dev/md/ds/d30). Before change: /dev/md/dsk/d200 swap no /dev/md/dsk/d300 /dev/md/rdsk/d30 /usr ufs 1 - no - After change: /dev/md/dsk/d20 swap no /dev/md/dsk/d30 /dev/md/rdsk/d300 /usr ufs 1 8.
Figure 3-20 Booting from the post-migration environment 11. When the host starts, confirm that the swap device and dump device have been correctly configured. Then, use the mount command to confirm that / (root) and /usr have been correctly configured.
# metainit -f d22 1 1 c6t50060E08000436D1d7s1 # metattach d20 d22 # metainit -f d32 1 1 c6t50060E08000436D1d7s6 # metattach d30 d32 13. Copy the SVM state database to the boot disk after migration.. # metadb -a c6t50060E08000436D1d6s7 # metadb -a c6t50060E08000436D1d7s7 14. Delete the SVM in the pre-migration environment. # metaclear d100 d101 d102 # metaclear d200 d201 d202 # metaclear d300 d301 d302 # metadb -d c8t50060E08000436D1d4s7 # metadb -d c8t50060E08000436D1d5s7 15.
Remove HDLM by referring to Removing HDLM on page 3-178. Checking the Path Configuration HDLM functions, such as load balancing and failover, are only available for HDLM management-target devices that have more than one active path. After you install HDLM or change the hardware configuration, check the structure and statuses of the paths. To check the path information, use the HDLM command's view operation. The following describes how to check path information by using the dlnkmgr command's view operation.
Setting Up HDLM Functions HDLM includes functions like the load balancing function, the automatic failback function, the error logging function, the audit logging function, etc. You can set up these functions by using the dlnkmgr command's set operation. The following subsections describe these setup methods.
Table 3-26 The default and recommended values for each function Function Default value Load-balancing Recommended value on on The Extended Least I/Os algorithm is used. The recommended algorithm depends on the operating environment. on on 30-minute check interval The recommended checking interval depends on the operating environment. on 1-minute check interval The recommended checking interval depends on the operating environment.
The following is an example command to set load-balancing: # /opt/DynamicLinkManager/bin/dlnkmgr set -lb on -lbtype exlio When load balancing is enabled by specifying on for the -lb parameter, specify one of the following algorithm values after the -lbtype parameter: • rr for the Round Robin algorithm • exrr for the Extended Round Robin algorithm • lio for the Least I/Os algorithm • exlio for the Extended Least I/Os algorithm • lbk for the Least Blocks algorithm • exlbk for the Extended Least Bl
If this condition is not satisfied, an error will occur and the warning message KAPL01080-W will be output. If this happens, change any of the following settings: the checking interval for automatic failbacks, the intermittent error-monitoring interval, or the number of times that the error needs to occur. If you set the number of times that the error needs to occur to 1, the above condition does not need to be satisfied.
To enable intermittent error monitoring, set it to on. Otherwise, set it to off. When on is specified, you can use the -intvl and -iemnum parameters to specify intermittent error conditions (the conditions used by the system to determine whether an intermittent error is occurring). Specify the monitoring interval for an intermittent error in the -intvl parameter, and the number of times that the error needs to occur in the -iemnum parameter.
The higher this value is set, the more information that will be output. As the amount of log information to be output increases, it will take less time to overwrite the old error log information with the new information. The following is an example of setting up the error log collection level by using a command: # /opt/DynamicLinkManager/bin/dlnkmgr set -ellv 2 Specify the error log collection level as a number. Setting the Trace Level You can set up the trace level for a trace file hdlmtrn.
You can specify a value (in kilobytes) from 100 to 2000000 for the error log file size. The specified value is applied for HDLM manager logs. If each error log file reaches the specified size, the information in the old error log file will be wrapped around and will be overwritten with new information. By specifying both the log file size and the number of log files, you can collect up to 32000000 kilobytes (approximately 30GB) of error logs in total.
# /opt/DynamicLinkManager/bin/dlnkmgr set -systfs 2000 Specify the size of the trace file in kilobytes. Setting the Number of Trace Files Trace files for which you can set the number of files are hdlmtrn.log (n indicates a file number from 1 to 64). For the number of the trace files, you can specify a value from 2 to 64.
Table 3-30 Values Indicating Audit Log Data Categories Value Explanation ss Audit log events of the StartStop category are collected. a Audit log events of the Authentication category are collected. ca Audit log events of the ConfigurationAccess category are collected. all Audit log events of the StartStop, Authentication, and ConfigurationAccess categories are all collected.
Value local7 or 23 Corresponding facility value in the /etc/syslog.conf file local7 This example shows how to specify the audit log facility: # /opt/DynamicLinkManager/bin/dlnkmgr set -audfac local0 Checking the Updated Settings This chapter describes how to check the updated settings by using the dlnkmgr command's set operation after settings have been changed. When you change some settings, you can display information about all of HDLM function settings.
(HNTRLib2) (/var/opt/hitachi/HNTRLib2/spool/hntr2n.log (n indicates a file number)). If a lot of integrated trace information is output, the older information might end up getting deleted in a very short amount of time. Also, if a large amount of integrated trace information is suddenly all output at the same time, any integrated trace information that is overflowing the buffer might not be saved into the integrated trace files.
Displaying the Hitachi Network Objectplaza Trace Library setup Menu To display the Hitachi Network Objectplaza Trace Library setup menu: 1. Log in as a user with root permissions. 2. Execute the following command: # /opt/hitachi/HNTRLib2/bin/hntr2utl2 The Hitachi Network Objectplaza Trace Library setup menu appears. If you do not want to change the settings, type e and then press the Enter key to quit the menu. The following explains how to modify each setting.
the Buffer Size Per Monitoring Interval Duration on page 3-146. The recommend value is 4096 for collecting an integrated trace. If you do not want to change the integrated trace file size, leave New Size(KB) blank, type !, and then press the Enter key to return to the Hitachi Network Objectplaza Trace Library setup menu. 3. Press the Enter key. The new setting is applied and the Hitachi Network Objectplaza Trace Library setup menu appears again.
2. Enter a larger size in New Size(KB). Set a new buffer size to fit the monitoring interval set in 5: Interval Timer. The default value is 64. You can specify a value from 8 to 2048 (KB). Set this to a value smaller than the integrated trace file size set in step 2 of Changing the Size of Integrated Trace Files on page 3-145. The recommended value is 256 for collecting an integrated trace.
A screen to set the monitoring interval for the number of messages output to the integrated trace file appears. The current value is displayed in Current Span (sec). 3-148 2. Enter a desired interval in New Span(sec). You can specify a value from 0 to 3600 (seconds). The default and recommended value is 0. If you do not want to change the monitoring interval, leave New Span(sec) blank, type !, and then press the Enter key to return to the Hitachi Network Objectplaza Trace Library setup menu.
If you do not want to change the maximum number of messages output, leave New Max(sec) blank, enter !, and then press the Enter key to return to the Hitachi Network Objectplaza Trace Library setup menu. 6. Press the Enter key. The new setting is applied and the Hitachi Network Objectplaza Trace Library setup menu appears again. Finishing the Hitachi Network Objectplaza Trace Library Settings To close the Hitachi Network Objectplaza Trace Library setup menu when you are finished: 1.
# /opt/hitachi/HNTRLib2/bin/hntr2mon -d & 7. Start the programs stopped in step 3. If you stopped programs other than HDLM in step 3, start them. Creating File Systems for HDLM (When Volume Management Software Is Not Used) When volume management software is not used, the file system directly uses HDLM devices. This section explains how to mount an HDLM-managed device as a UFS file system by specifying the name of the logical device file for an HDLM device.
Figure 3-21 An output example for the mount -p command Edit the /etc/vfstab file as appropriate based on the settings output. The shaded portion indicates the part of the /etc/vfstab file to set. The following figure shows an example of how to edit the /etc/vfstab file. Figure 3-22 An example of how to edit the /etc/vfstab file Add the shaded portion to the file. Setting Up VxVM When you are using VxVM, VxVM accesses an LU by using an HDLM device. VxVM will automatically recognize HDLM devices.
To register an HDLM device: 1. Make sure that VxVM recognizes HDLM devices. Execute the following commands sequentially to check whether VxVM recognizes HDLM devices. ¢ The dlnkmgr command's view -path operation Check the HDLM devices that must be recognized by VxVM. The following shows an example of executing the command: The following describes the procedure for checking whether the shaded HDLM devices are recognized. ¢ vxdisk command Make sure that the devices are recognized by VxVM.
Note The device name of the enclosure corresponding to the LU might vary before and after installing HDLM. Make sure that you perform step 1 to check the correspondence between the device name of the enclosure and the LU. If the vxdisk command displays the value of HDevName (cUtXdY) output by the dlnkmgr command's view -path operation, the HDLM device is recognized by VxVM. Go to step 2.
Note Do not add an HDLM device to a disk group registered as bootdg. Also, do not register a disk group containing an HDLM device in bootdg. 4. If you are using Sun Cluster, register the created disk group into a Sun Cluster disk device group. For details, see Linking VxVM and Sun Cluster on page 3-168. Creating VxVM Volumes A VxVM volume can be created by using the vxassist (or vxmake) command or Veritas Enterprise Administrator. For details, refer to the VxVM documentation.
Devices to Be Removed from VxVM The following figure shows devices that are recognized by VxVM before they are disabled. Figure 3-23 Devices That Are Recognized by VxVM before They Are Removed from VxVM You must remove the following non-HDLM devices that are recognized by VxVM. If the following devices are being managed by VxVM, HDLM functions, such as load balancing and failover, will not work normally.
# For logical device files of HDLM devices, the number of such files to be created is the same as the number of slices for one LU. The following figure shows the devices that are recognized by VxVM after the non-HDLM devices are disabled. Figure 3-24 Devices Recognized by VxVM after Removal of Devices from VxVM Only HDLM devices (c5t0d0) that are actually created are recognized. In this case, HDLM functions, such as load balancing and failover, work normally.
the configuration example shown in Figure 3-25 When HBA ports on the alternate path differ from those on the primary path connected to the Hitachi storage system on page 3-157. This figure shows a configuration that is recognized by VxVM before devices are removed. Figure 3-25 When HBA ports on the alternate path differ from those on the primary path connected to the Hitachi storage system To remove devices from VxVM, in the above configuration: 1.
The shaded portion, in the next step, is contained in the /etc/vx/ vxvm.exclude file. 2. Make an entry for the controller of the device you wish to remove from VxVM in the controllers section of the /etc/vx/vxvm.exclude file. In the /etc/vx/vxvm.exclude file, write the shaded portion of the execution result of the dlmvxexclude utility shown in step 1. The settings will take effect when the host is restarted. The following figure shows an example of the /etc/vx/vxvm.exclude file.
The devices displayed for HDevName (the shaded portion) are the HDLM devices that must be recognized by VxVM. In the above example, two devices, c5t0d0 and c5t0d1, must be recognized. 5. Execute the vxdisk command to list devices that are recognized by VxVM. The following shows an example of executing the command: If the HDLM devices output by the dlnkmgr command's view -path operation (the shaded portion) are displayed, the HDLM devices have been recognized by VxVM.
If any non-HDLM devices are ENABLED, perform the procedure described in Actions To Be Taken if an sd or ssd Device Has Not Been Suppressed from VxVM on page 3-164 of Removing Devices from VxVM on page 3-154 to set up vxvm.exclude again. When the display resulting from the vxdmpadm command is as shown in the following table, you do not need to reconfigure vxvm.exclude. Table 3-33 When setup of vxvm.exclude is unnecessary Display of the vxdmpadm command HDLM devices are not displayed.
Figure 3-27 When the HBA Port of the Physical Path Connected to a Hitachi Storage System Is the Same As the One Connected to a Third-party Storage System To remove devices from VxVM, in the above configuration: 1. Use the dlmvxexclude utility to acquire the mapping between controller numbers and devices. The devices to be removed from VxVM are output to the paths section in the execution result of the dlmvxexclude utility.
The shaded portion indicates the devices that need to be removed from VxVM. 2. Make an entry for the device you wish to removed from VxVM in the paths section of the /etc/vx/vxvm.exclude file. In the /etc/vx/vxvm.exclude file, write the shaded portion of the execution result of the dlmvxexclude utility shown in step 1. The settings will take effect when the host is restarted. The following figure shows an example of the /etc/vx/vxvm.exclude file. Figure 3-28 Example of a description in the /etc/vx/vxvm.
The devices displayed for HDevName (the shaded portion) are the HDLM devices that must be recognized by VxVM. In the above example, it is c4t0d0. 5. Execute the vxdisk command to list devices that are recognized by VxVM. The following shows an example of executing the command: If the HDLM devices output by the dlnkmgr command's view -path operation (the shaded portion) are displayed, the HDLM devices have been recognized by VxVM.
from VxVM on page 3-164 of Removing Devices from VxVM on page 3-154 to set up vxvm.exclude again. When the display resulting from the vxdmpadm command is as shown in the following table, you do not need to reconfigure vxvm.exclude. Table 3-34 When setup of vxvm.exclude is unnecessary Display of the vxdmpadm command HDLM devices are not displayed. Why suppression is unnecessary They are non-HDLM-managed devices such as a built-in disk. Only HDLM devices are displayed.
Example 1: The vxdisk command does not display HDLM devices but displays the sd or ssd devices corresponding to those HDLM devices. The vxdisk command displays the SCSI device c3t0d0, which corresponds to an HDLM device, instead of the HDLM device c5t0d0. Example 2: The vxdmpadm command outputs a non-HDLM device, and the STATE column shows ENABLED.
The vxdmpadm command displays c3t0d0 and c4t1d0 in addition to the HDLM device c5t0d0, and the STATE column shows ENABLED. If sd or ssd devices have not been removed normally from VxVM, as shown in Examples 1 and 2, take the following action: To suppress sd or ssd devices: 1. Execute the vxdmpadm command. In the following example, HDLM device c5t0d0 has been created and a non-HDLM device is ENABLED.
Figure 3-29 The /etc/vx/vxvm.exclude file (before editing) Figure 3-30 The /etc/vx/vxvm.exclude file (after editing) Among the non-HDLM devices in the ENABLED state, there might be two or more devices whose tXdY portions in the logical device file name (cUtXdYs2 format) are the same. In this case, change the currently assigned controller number in /etc/vx/vxvm.exclude. 5. After modifying all the sd or ssd device entries that have not been suppressed in /etc/vx/vxvm.
If the STATE column shows DISABLED, the device has been removed from VxVM. The following shows an example of executing the command: If there is more than one candidate for the controller number to be assigned in step 4, you might not be able to remove devices from VxVM in one operation. In this case, change the controller number you changed in step 4 to another controller number that was displayed using the vxdmpadm command, and then perform steps 5 and 6 again.
4. Verify that the VxVM disk group has been imported into only one of the nodes that constitute a cluster. If the VxVM disk group has been deported from all nodes, import the VxVM disk group to be registered in Sun Cluster into only one of the nodes. 5. If the VxVM disk group is not registered in Sun Cluster, execute the scsetup utility from the node into which you imported the VxVM disk group to be registered in Sun Cluster, and then register the VxVM disk group into a Sun Cluster disk device group.
KAPL05023-E Could not process the IOCTL(00004D03). (0000000D:00000000) 6. If the VxVM volume has not been created, create a VxVM volume in the disk group from the master node (the node that currently acts as the master for the disk device group). The following shows an example of executing the command: # vxassist -g dlmdg make vol-dg 10g 7.
• Even if you use a local diskset or shared diskset, do not specify the -r option when changing the path configuration and executing the dlmsetconf utility for creating the HDLM driver configuration definition file. When you specify the -r option, the minor number of the HDLM device used on the SDS will be changed, and you will no longer be able to use the metadevice. • Even if you use a local diskset or shared diskset, you cannot use the DiskSuite tool (metatool). Use the command line utility instead.
host1# metadb -a -f c1t1d0s1 c1t1d0s3 c1t1d0s7 2. Create a local metadevice. Use the HDLM device (slice) to create a local metadevice (logical disk). The types of metadevices include coupled, stripe, mirrored, RAID5, and transmeta devices. The following shows an example of executing the command to create local metadevice (d0) of type stripe (1) comprised of two HDLM devices (slices). host1# metainit d0 1 2 c5t1d0s7 c5t2d1s7 3. Create a file system on the local metadevice.
host1# metaset -s diskset1 -a -h host1 host2 3. Register a disk into the shared diskset. The following shows examples of executing the commands: When not using Sun Cluster: host1# metaset -s diskset1 -a c5t1d2 c5t1d3 When using Sun Cluster: host1# metaset -s diskset1 -a /dev/did/dsk/d10 /dev/did/dsk/ d11 4. Create a shared metadevice in the shared diskset.
Setting Up SVM If SVM is installed on an HDLM host, SVM uses the logical device file name of the HDLM device to access the HDLM-managed device. Therefore, after installing HDLM, you need to register the logical device file name of the HDLM device in SVM. When removing HDLM, unregister the logical device file name of the HDLM device and set the logical device file name of the sd or ssd device. Notes When using SVM, keep the following in mind.
Registering HDLM Devices If SVM is used, perform the following settings after installing HDLM, and register the SVM volume. To Use a Local Volume To use a local volume: 1. Create a replica of the state database. Create three replicas of the local state database on the disk (slice). Create these replicas on a disk that is not managed by HDLM, not shared, and has not been connected to a Fibre Channel. The following shows an example of executing the command: host1# metadb -a -f c1t1d0s1 c1t1d0s3 c1t1d0s7 2.
Create these replicas on all hosts that use the shared diskset. Make sure that all the following conditions are satisfied for the creation-target disk: ¢ The disk is not managed by HDLM. ¢ The disk is not shared. The disk is not connected via Fibre Channel. The following shows an example of executing the command: ¢ # metadb -a -f c1t1d0s1 c1t1d0s3 c1t1d0s7 2. Create a shared diskset. The following shows an example of executing the commands: host1# metaset -s diskset1 -a -h host1 host2 3.
host1# mkdir /diskset1 host1# mount /dev/md/diskset1/dsk/d1 /diskset1 When using Sun Cluster: host1# mkdir -p /global/diskset1/d1 host2# mkdir -p /global/diskset1/d1 host1# mount -g /dev/md/diskset1/dsk/d1 /global/diskset1/d1 Setting Up VCS When VCS is installed on an HDLM host, you can register HDLM devices, or VxVM disk groups containing HDLM devices, as resource groups. For details on registering resource groups, refer to the VCS documentation.
d: The total number of all HDLM devices #1 If the number of paths differs depending on an HDLM device, use the maximum number for the calculation. #2 The I/O timeout value used for sending a notification to the sd or ssd driver depends on the timeout settings for switches or HBAs. Removing HDLM This section explains how to return the HDLM environment to the way it was before HDLM was installed, and describes each step of the process.
Preparations for HDLM Removal Performing Operations on HDLM-Managed Devices To perform operations on HDLM-managed disks: 1. If necessary, back up the contents of the HDLM-managed devices to tape or some other storage medium. 2. For the HDLM devices, collect all relevant information on the correspondence between the logical device file names and the internal LUNs of the LUs in the storage system.
6. Use the following command to unmount all mount points that are using HDLM: # umount mount-point 7. If HDLM devices are being used in Solaris 10 ZFS, execute the following command to export all ZFS storage pools that are using HDLM: # zpool export pool-name 8. Edit the /etc/vfstab file.
As the utility instructs, enter the name of the VxVM disk device group you want to unregister. If the HDLM device has been configured as a Quorum device, perform the following operations to remove the Quorum device settings. 1. Check the device ID of the Quorum device. The following shows an example of executing the command: # /usr/cluster/bin/scconf -pv | grep Quorum # /usr/cluster/bin/scdidadm -L 2.
Make sure that the node that is in the maintenance mode has the following status: ¢ ¢ The node is in the Offline status. Both the current Quorum votes present and Quorum votes possible are 0. The following shows an example of executing the command: host2# /usr/cluster/bin/scstat -q 9. If the cluster is a two-node configuration, enter the install mode. If the cluster consists of three or more nodes, this step is not necessary so go to step 10.
For details on how to remove the VCS settings, refer to the VCS documentation. Remove VxVM Settings If you are using both HDLM and VxVM on the same host, before you remove HDLM, perform the following to delete the HDLM devices from VxVM. Note If you are using Sun Cluster, before deleting HDLM devices from VxVM, remove the settings for the disk device group (see Remove Sun Cluster Settings on page 3-180 of Preparations for HDLM Removal on page 3-179). For HDLM version 04-01 or earlier: 1.
When using a local metadevice 1. Execute the following command to unmount the file system The following shows an example of executing the command: host1# umount /mnt/filesystem1 2. Delete the local metadevice that is using the HDLM device. The following shows an example of executing the command to delete local metadevice d0. host1# metaclear d0 When using a shared diskset 1. Delete the shared metadevice. The following shows an example of executing the command.
The following shows an example of executing the command to delete SVM local volume d0. host1# metaclear d0 When using a shared diskset 1. Delete the shared volume. The following shows an example of executing the command. When not using Sun Cluster: host1# umount /diskset1 host1# metaclear -s diskset1 d1 When using Sun Cluster: host1# umount /global/diskset1/d1 host1# metaclear -s diskset1 d1 2. Delete the disk from the shared diskset. The following shows an example of executing the command.
¢ When removing HDLM in a Solaris 11 environment, be sure to execute /opt/DynamicLinkManager/bin/removehdlm. Do not use the pkg uninstall command to directly remove a Solaris package. To remove HDLM: 1. Log in to Solaris as a user with root permissions. 2. Execute the following command to remove HDLM: For Solaris 8, Solaris 9, or Solaris 10 # pkgrm DLManager For Solaris 11 # /opt/DynamicLinkManager/bin/removehdlm 3.
Then, execute the following command at the ok prompt on all nodes: boot -r Note While a node is starting, the message below may be output to the console or syslog: Could not stat: ../../devices/pseudo/dlmndrv@1/ dlmfdrv@0,0:c,raw path not loaded. This message is output because HDLM is removed and the HDLM devices are deleted.
A virtual disk backend is a location where virtual disk data is stored. A disk, disk slice, file, or volume (such as ZFS, SVM, and VxVM) can be used for the backend. In the control domain, register the install_disk file on a built-in disk as a virtual disk backend. The I/O domain recognizes this disk as a virtual disk. The device name of this virtual disk is /dev/[r]dsk/c0d0. In the I/O domain, register an HDLM device used in the storage system as a virtual disk backend.
To remove HDLM from a host in which LDoms have been set: 1. Log in to the control domain. 2. Execute the ldm stop command to stop the guest domain. For example, to stop a guest domain ldg2, execute the command as follows: # ldm stop ldg2 3. Execute the ldm unbind command to unbind the resource of the guest domain. For example, to unbind a resource of the guest domain ldg2, execute the command as follows: # ldm unbind ldg2 4.
For example, to remove virtual disks vdisk1 and vdisk2 from the guest domain ldg2, execute the command as follows: # ldm rm-vdisk vdisk1 ldg2 # ldm rm-vdisk vdisk2 ldg2 5. Execute the ldm rm-vdsdev command to stop exporting virtual disk backends from the service domain on the I/O domain to which HDLM will be installed.
9. On the control domain, execute the ldm add-vdisk command to assign the virtual disk backends to the guest domain. For example, to assign the virtual disk backends vol1 and vol2 exported from the service domain ldg1 to the guest domain ldg2, execute the command as follows: # ldm add-vdisk vdisk1 vol1@ldg1-vds0 ldg2 # ldm add-vdisk vdisk2 vol2@ldg1-vds0 ldg2 10. On the control domain, execute the ldm bind command to bind (connect) the resource of the guest domain ldg2.
Settings Needed After HDLM Removal VxVM Settings When using VxVM, execute the following command to confirm that the DEVICE column of the LUs previously managed by HDLM is displayed in the cWtXdYsZ format (for example, c6t0d10s2): # vxdisk list If the value is displayed in the cWtXdYsZ format, perform the following procedures: 1. Delete the /etc/vx/disk.info file. 2.
1. Remove the device ID information of devices that do not exist. The following shows an example of executing the command: # /usr/cluster/bin/scdidadm -C 2. Assign device IDs. The following shows an example of executing the command: # /usr/cluster/bin/scdidadm -r 3. Notify the server of the new device ID information. The following shows an example of executing the command: # /usr/cluster/bin/scdidadm -ui 4. Update the Sun Cluster global device namespace.
File System Settings For the file systems of the LUs that were managed by HDLM, specify settings that allow the Solaris sd or ssd driver to be mounted or specify settings that allow volume manager's volumes to be mounted after setup is performed again. After removing HDLM, the ZFS file system that was exported before the removal can be imported for use. Application Program Settings For the application programs that were using HDLM devices, specify the device file of the Solaris sd or ssd driver.
If you are unable to complete removal even though no programs other than HDLM are using HNTRLib2, contact your HDLM vendor or maintenance company if there is a maintenance contract for HDLM. Note If the log output directory set in HNTRLib2 was not the default directory, the log files will not be deleted during removal. In this case delete these files after removal.
¢ 3-196 Even if you attempt to remove HNTRLib2 when HNTRLib is installed, HNTRLib will not be removed. If no other programs are using HNTRLib, delete it manually.
4 HDLM Operation This chapter describes operating procedures for HDLM, including how to operate HDLM and the HDLM manager, and how to change the configuration of the operating environment.
Notes on Using HDLM This section explains the notes on using HDLM and using the environment in which HDLM is installed. Make sure that you read this section. Displaying Path Information • The path IDs displayed by using the dlnkmgr command's view operation differ depending on the order in which paths were detected when the host started. For this reason, you should always use the path name to specify a path.
The shaded portion indicates mapping from combinations of sd or ssd device types and instance numbers to HDLM device logical device file names. Storage System • Start up the storage system before you start up the hosts, so that Solaris can detect the storage system. • Do not change the vendor ID or product ID of a storage system. If you change either of them, HDLM will not be able to recognize the storage system.
If the host is started in single-user mode, the displayed path status indicates an error even when the actual path status is normal. Therefore, before you use HDLM devices in single-user mode, execute the procedure described in Initializing HDLM When the Host Is Started in Single-User Mode on page 4-4Operation in Single-User Mode on page 4-4. Upgrading Solaris In a local boot disk environment: When upgrading Solaris, first remove HDLM, and then upgrade Solaris.
3. Execute the following command to set the HDLM logical device file name in the HDLM filter driver: # /opt/DynamicLinkManager/bin/dlmsetdevname 4. Execute the following command to start the HDLM manager: # /etc/init.d/DLMManager start Tasks that Can Be Performed in Single-User Mode By executing the procedure described in Initializing HDLM When the Host Is Started in Single-User Mode on page 4-4, you can perform the following tasks in single-user mode while applications are inactive.
Maintenance Tasks on Devices Connected by Paths in the Boot Disk Environment Note the following points when performing maintenance on devices connected to the boot disk in a boot disk environment running Solaris: • Ensure that at least one path connected to the boot disk is online when you perform a maintenance task.
For details on the displayed items and their descriptions, see view (Displays Information) on page 6-32. Changing the Status of Paths This section explains how to change path statuses. Changing the Status of Paths to Online To change the status of paths to online: 1. Check the current status of the paths. To place paths online by specifying an HBA port, CHA port, or single path check the path name or PATH_ID used to manage the path.
KAPL01061-I 3 path(s) were successfully placed Online; 0 path(s) were not. Operation name = online # 3. Check to see if the statuses of all the applicable paths have changed. The following shows an example of executing the command: # /opt/DynamicLinkManager/bin/dlnkmgr view -path Changing the Status of Paths to Offline(C) To change the status of paths to Offline(C): 1. Check the current status of the paths.
# /opt/DynamicLinkManager/bin/dlnkmgr view -path Viewing LU Information This section explains how to display LU information by using the dlnkmgr command. To display LU information, execute the dlnkmgr command's view operation with the -lu parameter specified.
The following shows an example of executing the view operation: # /opt/DynamicLinkManager/bin/dlnkmgr view -drv PathID HDevName Device LDEV 000000 c6t5d0 sd74 9970/9980.15001.0670 000001 c6t5d1 sd75 9970/9980.15001.0671 000002 c6t5d2 sd76 9970/9980.15001.0672 000003 c6t5d3 sd77 9970/9980.15001.0673 000004 c6t5d4 sd78 9970/9980.15001.0674 000005 c6t5d5 sd79 9970/9980.15001.0675 000006 c6t5d0 sd80 9970/9980.15001.0670 000007 c6t5d1 sd81 9970/9980.15001.0671 000008 c6t5d2 sd82 9970/9980.15001.
# /opt/DynamicLinkManager/bin/dlnkmgr view -path Viewing and Setting Up the Operating Environment This section explains how to display and set up the HDLM operating environment. Viewing the Operating Environment To display the operating environment, execute the dlnkmgr command's view operation with the -sys and -sfunc parameters specified. The following shows an example of executing the command: # /opt/DynamicLinkManager/bin/dlnkmgr view -sys -sfunc HDLM Version : x.x.
• Load balancing • Path health checking • Automatic failback • Intermittent error monitoring • Dynamic I/O path control • Error log collection level • Trace level • Error log file size • The number of error log files • Trace file size • The number of trace files • Audit log data collection • Audit log facility • Number of times the same path can be used for load balancing • Number of times the same path can be used for extended load balancing For details on how to set up each f
view, completion time = yyyy/mm/dd hh:mm:ss # For details on the displayed items and their descriptions, see view (Displays Information) on page 6-32. Updating the License This section explains how to update the license. To update the license, execute the dlnkmgr command's set operation with the -lic parameter specified. When the confirmation message is displayed, enter y to execute, or n to cancel the command.
Number Of Elog Files : 2 Trace Level : 0 Trace File Size(KB) : 1000 Number Of Trace Files : 4 Path Health Checking : on(30) Auto Failback : on(1) Intermittent Error Monitor : off Dynamic I/O Path Control : off(10) HDLM Manager Ver WakeupTime Alive x.x.x-xx yyyy/mm/dd hh:mm:ss HDLM Alert Driver Ver WakeupTime ElogMem Size Alive x.x.x-xx yyyy/mm/dd hh:mm:ss 4096 HDLM Driver Ver WakeupTime Alive x.x.x-xx yyyy/mm/dd hh:mm:ss License Type Expiration Permanent KAPL01001-I The HDLM command completed normally.
Among the displayed items, HDLM Manager, HDLM Alert Driver, and HDLM Driver indicate the HDLM component information. Also, you can view information for each HDLM component. Execute the dlnkmgr command's view operation with the -sys and subsequent parameter specified.
The stop script that was set up during HDLM installation runs, and stops the HDLM manager. Execute the following dlnkmgr command's view operation to confirm that the HDLM manager has stopped. # /opt/DynamicLinkManager/bin/dlnkmgr view -sys -msrv HDLM Manager Ver WakeupTime Dead KAPL01001-I The HDLM command completed normally. Operation name = view, completion time = yyyy/mm/dd hh:mm:ss # When the HDLM Manager column shows Dead, the HDLM manager is inactive.
• Changing the LU default controller for a storage system • Adding, deleting, or changing a CHA port • Changing the topology of a storage system • Changing the Fibre Channel switches settings (including connection ports and zoning) • Changing a target ID • Changing the instance number of the HBA driver by changing the contents of the /etc/path_to_inst file • Switching the kernel mode • Inheriting logical device names during storage system migration Precautions Regarding Changes to the Config
• The LUN configuration was changed for the path to be used by HDLM to determine the logical device file name of the HDLM device. If the HDLM logical device file has been re-created by specifying the different file name, perform the following procedures: • Change the disk settings being used by the application programs.
obtained in step 1 (the driver name of the HDLM driver is dlmfdrv). Identify the path used to determine the HDLM logical device file name of the LU from the FilterPath property. # /usr/sbin/prtconf -vP The following shows an example of executing the command. This example shows the information on instance 1 of the HDLM driver (the driver name of the HDLM driver is dlmfdrv), which is contained in the output results of the prtconf command.
Dynamic Reconfiguration (DR) for Solaris In Solaris 8 or Solaris 9, you cannot use the Solaris dynamic reconfiguration (DR) functionality on HBAs that exist on a path managed by HDLM. To add, delete, or replace such HBAs, you need to restart the host after dynamic reconfiguration. Overview of Reconfiguring the HDLM Device Whether or not the host must be restarted after the reconfiguration depends on the contents of the reconfiguration.
The following operations are not supported in the dynamic reconfiguration of the HDLM device. To enable the following operations, restart the host with the reconfiguration option specified.
newly added path in HDLM, execute the dlmsetconf utility and then restart the host, with the reconfiguration option specified. • In Solaris 10 or Solaris 11, you can delete HDLM devices without restarting the host. However, if you cannot stop all of the programs that are using the HDLM devices that you want to delete or you cannot remove the mapping between the internal LUs and the storage port LUNs by performing operations on the storage system, you must restart the host to delete the HDLM devices.
Adding a New LU (When Not Using Sun Cluster) To add a new LU: 1. Set up the storage system and switch so that Solaris recognizes the LU.
3. Execute the dlmsetconf utility for creating the HDLM driver configuration definition file to update the HDLM driver configuration definition file (/ kernel/drv/dlmfdrv.conf). # /opt/DynamicLinkManager/bin/dlmsetconf 4. Set up LUs not to be managed by HDLM. If you do not want HDLM to manage one of the newly recognized LUs, set up the LU so that it is excluded from HDLM management. For details, See Using the HDLM Device Unconfiguration Function When Performing a New Installation of HDLM on page 3-72.
8. Register the added LU to the application. Adding a New LU By Restarting the Nodes (When Using Sun Cluster) Note Have the HDLM drivers on all nodes in the cluster recognize the LU newly recognized by the node, and make the configuration the same for each node. To add a new LU: 1. Set up the storage system and switch so that Solaris recognizes the LU. 2. Check the LU newly recognized by the node and its disk label.
Figure 4-2 Example of executing the format command In the above execution example, when you specify an LU, the message Disk not labeled. Label it now? will be output if a disk label has not been specified. In this case, enter y and specify a disk label. 3. If setup is performed for Quorum devices and disk device groups, on an LU newly recognized by the node, release these settings. When you unregister the SDS or SVM diskset, the diskset data will be deleted so first back up any necessary data.
If the key is displayed, perform the steps beginning with step 5. Even when you release the setup for Quorum devices and disk device groups, since the key used for disk reservation will remain registered, you will need to restart the node in non-cluster mode and unregister the key. If the key is not displayed, perform the steps beginning with step 9. 5. Simultaneously stop all nodes in the cluster. Execute the following command on any node: # /usr/cluster/bin/scshutdown -g0 -y 6.
When the dlmsetconf utility is executed, the KAPL10242-I message will be displayed prompting you to confirm the deletion of the logical device files for the sd or ssd device. If you respond by entering y, and then execute the dlmsetconf utility and restart the node, with the reconfiguration option specified, the logical device files of the sd or ssd device to be managed by HDLM will be deleted. If the logical device files are already backed up, enter y and execute the dlmsetconf utility.
message is not output next time the node starts. For details, refer to the Sun Cluster documentation. 14. Execute the dlnkmgr command's view operation on all nodes in the cluster, and display the status of each program. The following shows an example of executing the command: # /opt/DynamicLinkManager/bin/dlnkmgr view -sys HDLM Version : x.x.
17. Delete the device ID information for devices that do not exist. Execute the following command: # /usr/cluster/bin/scdidadm -C 18. Allocate the device ID. Execute the following command: # /usr/cluster/bin/scdidadm -r 19. Notify the node of the most recent device ID information. Execute the following command: # /usr/cluster/bin/scdidadm -ui 20. Update the Sun Cluster global device namespace.
# /usr/cluster/bin/scgdevs 21. When using, as a Quorum device, an HDLM device of an LU newly recognized for the first time by HDLM driver, set up the Quorum device on any node in the cluster. The following shows an example of executing the command: Check the device ID of the HDLM device.
procedure described in Adding a New LU By Restarting the Nodes (When Using Sun Cluster) on page 4-25. ¢ For details on the precautions for the dlmcfgmgr utility, see The dlmcfgmgr Utility for Managing the HDLM Configuration on page 7-16. To add a new LU: 1. Set up the storage system and switch so that Solaris recognizes the LU. 2. Check the LU newly recognized by the node and its disk label. Use the format command to check the LU and verify that the disk label is specified.
4. Make sure that the key used for disk reservation has been registered for the LU newly recognized by the node.
All sd or ssd devices connected to the host will be set as HDLM-managed devices. 7. If you do not want HDLM to manage one of the LUs newly recognized by the node, set up the LU so that it is excluded from HDLM management. For details, see Using the HDLM Device Unconfiguration Function When Performing a New Installation of HDLM on page 3-72. If you do not want to specify settings to exclude the LUs from HDLM management, skip to step 8. 8.
11. Delete the device ID information for devices that do not exist. Execute the following command: # /usr/cluster/bin/scdidadm -C 12. Allocate the device ID. Execute the following command: # /usr/cluster/bin/scdidadm -r 13. Notify the node of the most recent device ID information. Execute the following command: # /usr/cluster/bin/scdidadm -ui 14. Update the Sun Cluster global device namespace.
# /usr/cluster/bin/scgdevs 15. When using, as a Quorum device, an HDLM device of an LU newly recognized for the first time by HDLM driver, set up the Quorum device on any node in the cluster. The following shows an example of executing the command: Check the device ID of the HDLM device.
• Changing the host LUN • Changing, adding, or deleting a CHA port • Changing a target ID • Adding or replacing a Fibre Channel switch • Replacing, deleting, or adding an HBA # To change an LU, we recommend that you add a new LU and then delete the existing one, rather than changing a LU in the storage system management program. After adding the new LU, move data as required from the LU you are deleting to the new LU. For details on adding a new LU, see Adding a New Logical Unit on page 4-22.
If VxVM uses the LU you are deleting, delete the volumes and VxVM disk groups. To have another LU take over from the LU you are deleting, replace the LU. 8. If you are using SDS or SVM, perform the following as required: ¢ ¢ 9. If you are using a shared diskset, unregister the HDLM devices you are reconfiguring from the shared diskset. If the path used to reserve an LU is changed, it may not be possible to clear the reservation.
# /opt/DynamicLinkManager/bin/dlmsetconf The KAPL10242-I message appears when you execute the dlmsetconf utility. Type y in response. 15. Execute the following command to compare the HDLM driver unconfiguration definition file before and after execution of the dlmsetconf utility: # diff dlmfdrv.unconfbakn dlmfdrv.unconf Compare the old and new files to make sure that the reconfigured environment is recognized by HDLM. An example is shown below.
> ### Device=ssd83 target=134012 lun=3 > ### Path=/pci@8,700000/SUNW,emlxs@1/fp@0,0/ ssd@w50060e80033a9939,3 Execution example after a path is added by the addition of a CHA port: Check that a path for the same CHA port has been added: 26a31,32 > ### Device=ssd119 target=134012 lun=2 > ### Path=/pci@8,700000/SUNW,emlxs@1,1/fp@0,0/ ssd@w50060e80033a9939,2 28a35,36 > ### Device=ssd84 target=134012 lun=2 > ### Path=/pci@8,700000/SUNW,emlxs@1/fp@0,0/ ssd@w50060e80033a9939,2 34a43,44 > ### Device=ssd118 target=1
¢ boot -r (at the ok prompt) 20. Execute the dlnkmgr command's view operation specifying the -lu parameter, and check that the configuration has changed. 21. If you are using Sun Cluster, remove the device ID information of devices that do not exist. Execute the following command: # /usr/cluster/bin/scdidadm -C 22. If you are using Sun Cluster, assign device IDs. Execute the following command: # /usr/cluster/bin/scdidadm -r 23.
for details, see To Use a Local Metadevice on page 3-171 in Registering HDLM Devices on page 3-171. 28. Recreate any volumes you deleted in step 8. For details, see To Use a Local Volume on page 3-175 in Registering HDLM Devices on page 3-175. 29. If you unregistered VxVM disk groups from the resource groups in step 6, re-register the resource groups, referring to the VCS documentation. 30. If you recreated a file system, volume, or metadevice, restore the data backed up in step 1. 31.
4. Execute the dlmcfgmgr utility for managing the HDLM configuration to delete the LU, which is disconnected from the host, from the HDLM driver. # /opt/DynamicLinkManager/bin/dlmcfgmgr -a The HDLM device will be deleted without restarting the host. 5. Execute the dlnkmgr command's view -path operation to check the path status after the configuration change.
# cfgadm -al Ap_Id Occupant system:slot0 configured system:slot1 configured 2. Type Cond cpu/mem ok sbus-upa ok Receptacle connected connected Check the newly recognized path in the syslog. The following is an example of the command. Specify the number of lines to display taking into consideration that three to four lines for each recognized path are output to the log. # tail -10 /var/adm/messages Apr 2 16:02:57 shiga genunix: [ID 408114 kern.
# /opt/DynamicLinkManager/bin/dlmcfgmgr -a KAPL10373-I This operation will change the configuration of HDLM devices, and delete device files of sd/ssd devices managed by HDLM. Do you want to continue? [y/n]:y KAPL10771-I The path (FilterPath = /pci@400/pci@2/pci@0/pci@8/ SUNW,qlc@0,1/fp@0,0/ssd@w50060e8010027a84,5) was successfully added. KAPL10771-I The path (FilterPath = /pci@400/pci@2/pci@0/pci@8/ SUNW,qlc@0/fp@0,0/ssd@w50060e8010027a84,5) was successfully added.
ssd@w50060e8010027a80,4 ssd4123 000050 Online /devices/ pci@400/pci@2/pci@0/pci@8/SUNW,qlc@0,1/fp@0,0/ ssd@w50060e8010027a84,4 ssd4122 000051 Online /devices/ pci@400/pci@2/pci@0/pci@8/SUNW,qlc@0/fp@0,0/ ssd@w50060e8010027a84,4 0003 c7t50060E8010027A80d5 ssd4125 000052 Online /devices/ pci@400/pci@2/pci@0/pci@8/SUNW,qlc@0,1/fp@0,0/ ssd@w50060e8010027a80,5 ssd4124 000053 Online /devices/ pci@400/pci@2/pci@0/pci@8/SUNW,qlc@0/fp@0,0/ ssd@w50060e8010027a80,5 ssd4127 000054 Online /devices/ pci@400/pci@2/pci@0/
device, therefore you should also see The dlmsetconf Utility for Creating the HDLM Driver Configuration Definition File on page 7-23. Changing an HDLM-managed Device to a Non-HDLM-Managed Device To change an HDLM-managed device to a non-HDLM-managed device: 1. Check the LDEV information of the LU to be removed from HDLM management. Use the following command to check the LDEV information: # /opt/DynamicLinkManager/bin/dlnkmgr view -drv PathID HDevName Device LDEV 000000 c7t0d0 sd46 9970/9980.15001.
Figure 4-5 Example of executing the format command In this example, the shaded portion indicates the path information of the LU to be managed by HDLM. In the above execution example, when you specify an LU, the message Disk not labeled. Label it now? will be output if a disk label has not been specified. In this case, enter y and specify a disk label. 2. Edit the HDLM driver unconfiguration definition file (/opt/ DynamicLinkManager/config/dlmfdrv.unconf).
Figure 4-6 The /opt/DynamicLinkManager/config/dlmfdrv.unconf file (before editing) Figure 4-7 The /opt/DynamicLinkManager/config/dlmfdrv.unconf file(after editing) The shaded portion indicate the text to be edited. Note Do not edit anything other than adding the hash mark (#) to the beginning of the LDEV property line. 3. Apply the contents that were edited in step 2 to the HDLM driver configuration definition file (/kernel/drv/dlmfdrv.conf).
Specifying Whether a Logical Unit Is To Be Managed by HDLM (When Using Sun Cluster) Changing an HDLM-Managed Device to a Non-HDLM-Managed Device Note ¢ ¢ When you remove an HDLM-managed LU from management, perform the remove operations on all nodes in the cluster, and make the configuration the same on each node. When you remove an HDLM-managed LU from management, restart the host. To change an HDLM-managed device to a non-HDLM-managed device: 1. Log in to Solaris as a user with root permissions. 2.
specifies that the LU indicated by that LDEV attribute is not to be managed by HDLM. The following figure shows examples of editing the HDLM driver unconfiguration definition file (/opt/DynamicLinkManager/config/ dlmfdrv.unconf). Figure 4-8 The /opt/DynamicLinkManager/config/dlmfdrv.unconf file (before editing) Figure 4-9 The /opt/DynamicLinkManager/config/dlmfdrv.unconf file (after editing) The shaded portion indicates the text to be edited.
Figure 4-10 Result output by executing the dlmvxexclude utility Figure 4-11 Coding in the /etc/vx/vxvm.exclude file The shaded portion indicates the device not to be managed by HDLM and to be released from the VxVM disabled state. For details on the dlmvxexclude utility, see The dlmvxexclude Utility for Assisting Creation of the VxVM Configuration File on page 7-30. 7. Stop all nodes in the cluster. On the desired one node, execute the following command: # /usr/cluster/bin/scshutdown -g0 -y 8.
Note The following message might be output to the console window or syslog while the node is running: Unable to open '/dev/rdsk/c5t50060E80033A9902d2s2' This message is output when an HDLM-managed device has been changed to a non-HDLM-managed device. Perform the procedure for updating the device ID information (steps 12 to 15) so that this message is not output next time the node starts. For details, refer to the Sun Cluster documentation. 9.
You can check the LU by using the format command. The following figure shows an example of the format command. The shaded portion represents an LU not managed by HDLM. Figure 4-12 Example of executing the format command 12. Delete the device ID information for devices that do not exist. Execute the following command: # /usr/cluster/bin/scdidadm -C 13. Allocate the device ID.
16. When using a non-HDLM-managed device, as a Quorum device, set up the Quorum device in any node on the cluster. The following shows examples of executing the commands: Check the device ID of the device not to be managed by HDLM: # /usr/cluster/bin/scdidadm -L Set up the Quorum device: # /usr/cluster/bin/scconf -a -q globaldev=d4 Execute the following command to clear the install mode of a cluster when it consists of two nodes: # /usr/cluster/bin/scconf -c -q reset 17.
Figure 4-13 Example of executing the format command In the above execution example, when you specify an LU, the message Disk not labeled. Label it now? will be output if a disk label has not been specified. In this case, enter y and specify a disk label. 3. If setup is being performed for Quorum devices and disk device groups in the LU to be managed by HDLM, release the LU settings. When you unregister the SDS or SVM diskset, the diskset data will be deleted so first back up any necessary data.
Execute the following command on any node: # /usr/cluster/bin/scshutdown -g0 -y 6. Start all nodes in the cluster in non-cluster mode. At the ok prompt, execute the boot -x command. 7. Unregister the key for the LU to be managed by HDLM.
Figure 4-15 The /opt/DynamicLinkManager/config/dlmfdrv.unconf file (after editing) The shaded portion indicates the text to be edited. Note Do not make any modifications other than adding or deleting the hash mark (#) in front of the LDEV property line corresponding to the LU. 10. Apply the settings in the HDLM driver unconfiguration definition file (/ opt/DynamicLinkManager/config/dlmfdrv.unconf) to the HDLM driver configuration definition file (/kernel/drv/dlmfdrv.
message is not output next time the node starts. For details, refer to the Sun Cluster documentation. 14. Execute the dlnkmgr command's view operation on all nodes in the cluster, and display the status of each program. The following shows an example of executing the command: # /opt/DynamicLinkManager/bin/dlnkmgr view -sys HDLM Version : x.x.
17. Delete the device ID information for devices that do not exist. Execute the following command: # /usr/cluster/bin/scdidadm -C 18. Allocate the device ID. Execute the following command: # /usr/cluster/bin/scdidadm -r 19. Notify the node of the most recent device ID information. Execute the following command: # /usr/cluster/bin/scdidadm -ui 20. Update the Sun Cluster global device namespace. Execute the following command: # /usr/cluster/bin/scgdevs 21.
Set up the Quorum device: # /usr/cluster/bin/scconf -a -q globaldev=d4 Execute the following command to clear the install mode of a cluster when it consists of two nodes: # /usr/cluster/bin/scconf -c -q reset 22.
You can check the LU and verify that the disk label is specified by using the format command. Figure 4-16 Example of executing the format command on page 4-62 shows an example of executing the format command. The shaded portion represents the LU to be managed by HDLM. Figure 4-16 Example of executing the format command In the above execution example, when you specify an LU, the message Disk not labeled. Label it now? will be output if a disk label has not been specified.
If the key is displayed, place the LU under HDLM management by following the steps from step 5 of Changing a Non-HDLM-Managed Device to an HDLM-Managed Device (When the Node Must Be Restarted) on page 4-55 in Specifying Whether a Logical Unit Is To Be Managed by HDLM (When Using Sun Cluster) on page 4-50.
6. Apply the settings in the HDLM driver unconfiguration definition file (/ opt/DynamicLinkManager/config/dlmfdrv.unconf) to the HDLM driver configuration definition file (/kernel/drv/dlmfdrv.conf): Execute the following command: # /opt/DynamicLinkManager/bin/dlmsetconf -u 7. Execute the dlmcfgmgr utility shown below to make the HDLM driver recognize the LU to be managed by HDLM: # /opt/DynamicLinkManager/bin/dlmcfgmgr -a 8. Check the status of the path accessing the LU to be managed by HDLM.
# /usr/cluster/bin/scdidadm -ui 12. Update the Sun Cluster global device namespace. Execute the following command: # /usr/cluster/bin/scgdevs 13. When using an HDLM-managed device, as a Quorum device, set the Quorum device on any node in the cluster.
Inheriting logical device names during storage system migration This subsection describes inheritance of logical device names during storage system migration. Note that if the volume manager is either SVM or VxVM, logical device names cannot be inherited during storage system migration.
switches, cables between hosts and the storage system devices, and the HBA driver. 6. Set up switches and other devices so that the logical units in the destination storage system are recognized by the OS. # cfgadm -c configure ap_id For the value to be specified for ap_id, see the listing of hardware information displayed by the cfgadm -al command. The following shows an example of executing the command. # cfgadm -al Ap_Id Occupant system:slot0 configured system:slot1 configured 7.
Then, execute the dlmsetconf utility with the -u parameter specified to update configuration definition files. # /opt/DynamicLinkManager/bin/dlmsetconf -u 11. Execute the dlmcfgmgr utility as shown below so that HDLM recognizes the added logical units. # /opt/DynamicLinkManager/bin/dlmcfgmgr -a 12. Check the status of the paths to the added logical units. # /opt/DynamicLinkManager/bin/dlnkmgr view -path 13. Configure the upper-level software for the added logical units and other items.
5 Troubleshooting This chapter describes how to properly use HDLM error information, and then how to resolve the problem, if an error occurs in HDLM. The resolutions for path errors, HDLM program errors, and other types of errors are all described separately. If you need technical support, see Getting help on page xvii.
Information collected by using the DLMgetras utility for collecting HDLM error information Immediately after an error occurs, execute the DLMgetras utility to collect HDLM error information, because restarting the machine may delete existing error information already collected by the DLMgetras utility. For details on the DLMgetras utility and the error information it collects, see The DLMgetras Utility for Collecting HDLM Error Information on page 7-3.
The path name indicates a physical path. When you modify the system configuration or replace a piece of hardware, you should check the path names to identify the physical paths that will be affected by the change. A path name consists of the following four elements, separated by periods: ¢ Host port number (hexadecimal) ¢ Bus number (hexadecimal) ¢ Target ID (hexadecimal) Host LU number (hexadecimal) The path name is also the same as the PathName displayed by the dlnkmgr command's view operation.
Figure 5-1 Error Location When the KAPL08022-E Message is Output The following figure shows the troubleshooting procedure when the KAPL08022-E message is output. Figure 5-2 Troubleshooting Procedure When a Path Error Occurs If an error occurs in a boot disk environment on a path connected to the boot disk, perform the procedure described in Actions to Take for a Path Error in a Boot Disk Environment on page 5-6.
The following shows the procedure for using the dlnkmgr command to handle a path error. Examining the Messages Examine the message that is output to syslog in the host by using applications or tools for monitoring messages. If the KAPL08022-E message is output, view the message to check the path in which the error has occurred. For details on the message, see Checking Error Information in Messages on page 5-2. Obtain Path Information Obtain path information.
If any path cannot be placed online due to an error, the KAPL01039-W message will appear. To ignore such paths and to continue processing, type y. To cancel the processing, type n. Recheck the statuses of the paths that cannot be placed online, and resolve the problem. Actions to Take for a Path Error in a Boot Disk Environment This section describes the action to take if an error occurs in a boot disk environment on a path connecting to the boot disk.
¢ The message output to the console by Solaris ¢ The message output to the console by the HBA driver If an error occurs on one or more of the paths to the boot disk HDLM outputs the KAPL08022-E message to the console, syslog, and error log after the HDLM manager has started.
Figure 5-3 Troubleshooting Procedure When a Program Error Occurs The following shows the procedure for handling a program error by using the dlnkmgr command. Examining the Messages Examine the messages that are output to the host syslog. If an error occurs in an HDLM program, a message other than KAPL08xxx is output to syslog. Check the output message. Messages with error level E (Error) or higher require corrective action.
# /opt/DynamicLinkManager/bin/dlnkmgr view -sys If the KAPL01012-E message appears as a result of executing the command The following shows the KAPL01012-E message: KAPL01012-E Could not connect the HDLM manager. Operation name = view If this message appears, start the HDLM manager. For details on how to start the HDLM manager, see Starting the HDLM Manager on page 4-15.
5-10 Troubleshooting Hitachi Dynamic Link Manager User Guide for Solaris®
6 Command Reference This chapter describes the HDLM command (dlnkmgr) and its operations.
Overview of the HDLM Command dlnkmgr This section describes how to specify the HDLM command dlnkmgr and its subcommands (called operations in HDLM). Command format Enter the command using the following format: dlnkmgr operation [parameter [parameter-value]] dlnkmgr The command name. operation The type of operation entered after dlnkmgr. parameter A value required for an operation. parameter-value A value required for an operation parameter.
Note: ¢ ¢ Execute the command as a user with root permissions. To specify a value that contains a space in its parameter, enclose the entire value in double quotation marks ("). clear (Returns the Path Statistics to the Initial Value) The dlnkmgr command's clear operation clears the statistics (I/O count and I/O errors) of all paths that are managed by the HDLM, and returns them to their initial value.
KAPL01001-I The HDLM command completed normally. Operation name = clear, completion time = yyyy/mm/dd hh:mm:ss # To display the format of the clear operation -help Displays the format of the clear operation. Example # /opt/DynamicLinkManager/bin/dlnkmgr clear -help clear: Format dlnkmgr clear -pdst [-s] KAPL01001-I The HDLM command completed normally.
If you do not specify any operations, the help operation displays all operations available for the dlnkmgr command. Examples Example 1 The following example shows how to display all the operations available for the dlnkmgr command. # /opt/DynamicLinkManager/bin/dlnkmgr help dlnkmgr: Format dlnkmgr { clear | help | offline | online | set | view | monitor | add} KAPL01001-I The HDLM command completed normally.
monitor | add } KAPL01001-I The HDLM command completed normally. Operation name = help, completion time = yyyy/mm/dd hh:mm:ss # Example 3 The following example shows how to display the operations that can be specified for the help operation # /opt/DynamicLinkManager/bin/dlnkmgr help help help: Format dlnkmgr help { clear | offline | online | set | view | monitor | add} KAPL01001-I The HDLM command completed normally.
To display the format of the offline operation /opt/DynamicLinkManager/bin/dlnkmgr offline -help Parameters To place paths offline -path Indicates that the target of the operation is a path managed by HDLM. This parameter is optional because offline is always used for paths, so it is assumed. Make sure that you specify the paths to be taken offline by using the hba, -hbaid, -cha, -chaid, -pathid, or -device parameter. -hba host-port-number.
in Parameters on page 6-34. Example The following example shows how to place offline all paths connected to the HBA port whose HBA-port-ID is 00001. # /opt/DynamicLinkManager/bin/dlnkmgr offline -hbaid 00001 KAPL01102-I All the paths which pass the specified HBA port will be changed to the Offline(C) status. Is this OK? [y/n]:y KAPL01103-I If you are sure that there would be no problem when all the paths which pass the specified HBA port are placed in the Offline(C) status, enter y. Otherwise, enter n.
Use this parameter to place offline, at one time, all paths that pass through a specific CHA port. The command will place offline all paths connected to the CHA port that has the specified CHA-port-ID. To display the CHA port IDs, execute the following view operation: # /opt/DynamicLinkManager/bin/dlnkmgr view -cha For details about how to execute the view operation to display the CHA port IDs, see To display CHA port information on page 6-65 in Parameters on page 6-34.
Example The following shows an example of how to place offline all paths whose SCSI device name is sd52 while confirming command operation. # /opt/DynamicLinkManager/bin/dlnkmgr offline -path -device sd52 KAPL01052-I The currently selected paths will be changed to the Offline(C) status. Is this OK? [y/n]:y KAPL01053-I If you are sure that there would be no problem when the path is placed in the Offline(C) status, enter y. Otherwise, enter n.
AutoPATH_ID { 000000 - 999999 }(Decimal) HBA_ID { 00000 - 99999 }(Decimal) CHA_ID { 00000 - 99999 }(Decimal) KAPL01001-I The HDLM command completed normally. Operation name = offline, completion time = yyyy/mm/dd hh:mm:ss # Reference Executing the view operation together with standard UNIX commands enables you to filter the path information listed for a specific HBA port or CHA port. For details on the view operation, see view (Displays Information) on page 6-32.
[-path] [-hba host-port-number.bus-number |-hbaid HBA-port-ID |-cha -pathid AutoPATH_ID |-chaid CHA-port-ID |-pathid AutoPATH_ID |-device SCSI-device-name |-hapath [-lu -pathid AutoPATH_ID] |-dfha [-lu -pathid AutoPATH_ID]] [-s] To display the format of the online operation /opt/DynamicLinkManager/bin/dlnkmgr online -help Parameters To place paths online -path Indicates that the target of the operation is a path managed by HDLM.
When the confirmation message is displayed, the user enters y to continue, or n to cancel the operation. # /opt/DynamicLinkManager/bin/dlnkmgr online -hba 0001.0000 KAPL01057-I All the paths which pass the specified HBA will be changed to the Online status. Is this OK? [y/n]:y KAPL01061-I 3 path(s) were successfully placed Online; 0 path(s) were not. Operation name = online # -hbaid HBA-port-ID Use this parameter to place online, at one time, all paths that pass through a specific HBA port.
the view operation to find the current AutoPATH_ID of the target path, before executing the online operation. Example The following example shows how to place online the paths connected to the CHA port 0A. In this example, a path whose AutoPATH_ID is 000002 is connected to the target CHA port: # /opt/DynamicLinkManager/bin/dlnkmgr online -cha -pathid 000002 KAPL01057-I All the paths which pass the specified CHA port will be changed to the Online status.
Use this parameter to place online all paths connected to the specified SCSI device. For SCSI-device-name, specify the value of Device displayed by executing the view operation as follows: # /opt/DynamicLinkManager/bin/dlnkmgr view -lu For details on how to execute the view operation and display the SCSI device name, see To display LU information on page 6-53 in Parameters on page 6-34. Only one case-sensitive parameter value can be specified.
if you want to skip the response to the confirmation message: for example, when you want to execute the command in a shell script or batch file. Example The following example shows how to place a path, whose AutoPATH_ID is 000002, online without asking for confirmation of command execution from the user: # /opt/DynamicLinkManager/bin/dlnkmgr online -pathid 2 -s KAPL01061-I 1 path(s) were successfully placed Online; 0 path(s) were not.
We recommend that you execute the following command and verify the information on the target paths before you execute the online operation to place online all the paths connected to a specific HBA port or CHA port. Example 1 The following example shows how to filter and display the information on all paths that pass through the HBA port whose host port number is 0004 and bus number is 0000. # /opt/DynamicLinkManager/bin/dlnkmgr view -path | grep 0004.
|-audlog {on [-audlv audit-log-data-collection-level] [-category [[ss] [a] [ca]|all]]|off} |-audfac facility-value |-lbpathusetimes number-of-path-use-times |-expathusetimes number-of-path-use-times |-dpc {on|off} [-pathid path-ID -lu|-pathid path-ID -storage] |-dpcintvl checking-interval } [-s] To display the format of the set operation /opt/DynamicLinkManager/bin/dlnkmgr set -help Parameters To set up the HDLM operating environment The table below shows the default and recommended values for each settin
Item name Intermittent error monitoring Default value off Recommended value off Collecting audit log off data The recommended value depends on the operating environment. Set on, if you want to collect audit log data. Audit log facility user local0 to local7 Number of times the same path can be used for load balancing 1 The recommended value depends on the operating environment.
The following shows the log files in which an error log collection level can be set: HDLM manager logs: dlmmgrn.log (n indicates a file number from 1 to 16) Table 6-3 Values of the Error Log Collection Level on page 6-20 shows the values of the error log collection level. Table 6-3 Values of the Error Log Collection Level Value Description 0 Collects no error log. 1 Collects error information at the Error level or higher. 2 Collects error information at the Warning level or higher.
The target log files for which the number of files can be set are HDLM manager logs (dlmmgrn.log (n indicates a file number from 1 to 16)) only. -systflv trace-level Specify the trace output level. The trace files for which trace levels can be set are hdlmtrn.log (n indicates a file number from 1 to 64). Table 6-4 Trace Level Values on page 6-21 shows the values of the trace level. When an error occurs, we recommend that you set the value of the trace level to 1 or higher and re-collect log information.
The trace files for which the number of files can be set are hdlmtrn.log (n indicates a file number from 1 to 64). -pchk {on [-intvl checking-interval]|off} Enables or disables path health checking. on: Enabled off: Disabled For a standby host, or a host connected to the Thunder 9500V series, or Hitachi AMS/WMS series storage system, we recommend that you activate path health checking so that you can detect errors in paths where I/Os operations do not occur.
¢ Paths where an error occurred and for which the KAPL08022-E message was displayed. Path where an error occurred during the startup of the HDLM manager. To prevent intermittent errors from deteriorating I/O performance, we recommend that you also enable intermittent error monitoring when the automatic failback function is enabled. Intermittent error monitoring is specifiable only when automatic failbacks are enabled.
Therefore, if you re-enable automatic failbacks and do not change the interval, path status checks will be executed at the interval already stored in the system. -iem { on [-intvl error-monitoring-interval] [-iemnum number-of-timeserror-is-to-occur] | off } Enables or disables intermittent error monitoring. on: Enabled off: Disabled Intermittent error monitoring can be set when automatic failback is set to on.
-intvl error-monitoring-interval Specify the monitoring interval for an intermittent error. Use a value from 1 to 1440 minutes. The default is 30. During intermittent error monitoring, if changes are made in the settings of the intermittent error monitoring interval, the error count and the elapsed time measured since monitoring starts are set to 0, and monitoring starts by using the new settings.
The executable operations for the automatic failback function and intermittent error monitoring depend on the settings for those functions. The table below shows the relationship between the settings and available operations for automatic failback and intermittent error monitoring. Table 6-5 Relationship Between the Setting for the Automatic Failback Function and Intermittent Error Monitoring and the Executable Operations Setting AFB on IEM on Executable operation Set AFB to on.
Setting AFB off IEM on#2 off Executable operation Result of operation Set AFB to on. AFB and IEM are enabled.#1 Set AFB to off. The operations of AFB and IEM do not change. Set AFB to on. AFB is enabled. Set AFB to off. The operations of AFB and IEM do not change. Legend: AFB: Automatic failback IEM: Intermittent error monitoring #1 When this condition is not satisfied, the KAPL01080-W message is output and an error occurs. The status of intermittent error monitoring does not change.
A message confirming that the license key has been registered is displayed, depending on the license key type described in the license key file. For a temporary license key or emergency license key, the expiration period is also displayed (KAPL01071-I, KAPL01072-I). The following table lists and describes the license key types. Table 6-6 License Key Types Type Description Permanent license key Permanent license keys are valid for using HDLM permanently.
-audlv audit-log-data-collection-level Specifies the severity level of audit log data to be collected. The table below lists and describes the values used for this setting. The default is 6. Table 6-7 Values Indicating Audit Log Data Collection Levels Value (severity) 0 Explanation No audit log data is collected. 1 2 Critical-level audit log data is collected. 3 Critical-level and Error-level audit log data is collected. 4 Critical-level, Error-level, and Warning-level audit log data is collected.
Value Corresponding facility value in the /etc/syslog.conf file local1 or 17 local1 local2 or 18 local2 local3 or 19 local3 local4 or 20 local4 local5 or 21 local5 local6 or 22 local6 local7 or 23 local7 -lbpathusetimes number-of-path-use-times Specifies the number of times the same path can be used for I/O operations when the Round Robin (rr), Least I/Os (lio), or Least Blocks (lbk) algorithm is used for load balancing. You can specify a decimal (base 10) value from 0 to 999999.
Specifies the checking interval (in minutes) for reviewing information about switching of controllers performed by the storage system which is used in the dynamic I/O path control function. Specify a value in the range from 1 to 1440. The default value is "10". -s Executes the command without displaying the message asking for confirmation of command execution from the user. Specify this parameter if you want to skip the response to the confirmation message.
Number-Of-ElogFiles Value 2) TraceLevel Value 0) TraceFileSize Value 1000) Number-Of-TraceFiles Value 4) Interval-Time Value 30) (pchk) Interval-Time Value 1) (afb) Error-Monitor-Interval Value 30) Number-Of-Times Value 3) AudlogLevel Value 6) Category-Value Value all) Facility-Name { 2 - 16 }(Files) (Default { 0 | 1 | 2 | 3 | 4 } (Default { 100 - 16000 }(KB) (Default { 2 - 64 }(Files) (Default { 1 - 1440 }(Minute) (Default { 1 - 1440 }(Minute) (Default { 1 - 1440 }(Minute) (Default { 1 - 99 }(
[-t] To display path information To display path information /opt/DynamicLinkManager/bin/dlnkmgr view -path [-hdev host-device-name] [-stname] [-iem] [-srt {pn|lu|cp}] [-hbaportwwn] [-t] To display path information (by selecting a display item) /opt/DynamicLinkManager/bin/dlnkmgr view -path -item [pn] [dn] [lu] [cp] [type] [ic] [ie] [dnu] [hd] [iep] [hbaportwwn] [vid] [ha] [-hdev host-device-name] [-stname] [-srt {pn|lu|cp}] [-t] To display a summary of path information /opt/DynamicLinkManager/bin/dlnkmg
To display a summary of LU information /opt/DynamicLinkManager/bin/dlnkmgr view -lu -c [-t] To display a summary of LU information (by adding items to be displayed) /opt/DynamicLinkManager/bin/dlnkmgr view -lu -c -item [slpr] [-t] To display HBA port information /opt/DynamicLinkManager/bin/dlnkmgr view -hba [-srt pb] [-c|portwwn] [-t] To display CHA port information /opt/DynamicLinkManager/bin/dlnkmgr view -cha [-srt cp] [-t] To display corresponding information about an HDLM device, sd or ssd device, a
Use one of the sub-parameters (following -sys) to specify the program information that you want to display. If you do not specify a subparameter, the command displays all of the program information except the information about the audit log data collection settings, the number of times the same path can be used for load balancing, and the number of times the same path can be used for extended load balancing.
Parameter and program information to be displayed Item Description • 2: Collects error information at the Warning level or higher. • 3: Collects error information at the Information level or higher. • 4: Collects error information at the Information level or higher (including maintenance information). Elog File Size Size of the error log file in kilobytes (KB) Number Of Elog Number of error log files Files Trace Level Trace output level: • 0: Does not output any trace.
Parameter and program information to be displayed Item Description When automatic failback is off, intermittent error monitoring is disabled although Intermittent Error Monitor will be shown as on. When the automatic failback function is on, intermittent error monitoring will be enabled.
Parameter and program information to be displayed Information about the HDLM driver -lic Item Description Dead: Stopped Ver Version number of the HDLM driver WakeupTime Startup time of the HDLM driver License Type License type Information about the HDLM license Expiration • Permanent • Temporary • Emergency License expiration: • When using a permanent license: - • When using a temporary license or emergency license: The license expiration period is displayed in the following format: yyyy
Parameter and program information to be displayed Item Audit Log Facility Description The name of the audit log facility is displayed. user, local0, local1, local2, local3, local4, local5, local6, or local7 is displayed. If the collection of audit log data is disabled, a hyphen (-) is displayed.
Intermittent Error Monitor : off Dynamic I/O Path Control : off(10) KAPL01001-I The HDLM command completed normally. Operation name = view, completion time = yyyy/mm/dd hh:mm:ss # Example 2 The following example shows how to display information about the HDLM manager: # /opt/DynamicLinkManager/bin/dlnkmgr view -sys -msrv HDLM Manager Ver WakeupTime Alive x.x.x-xx yyyy/mm/dd hh:mm:ss KAPL01001-I The HDLM command completed normally.
Audit Log Category : Audit Log Facility : KAPL01001-I The HDLM command completed normally. Operation name = view, completion time = yyyy/mm/dd hh:mm:ss # Example 7 The following example shows how to display the number of times the same path can be used for load balancing: # /opt/DynamicLinkManager/bin/dlnkmgr view -sys -lbpathusetimes Times Same Path Was Used : 1 KAPL01001-I The HDLM command completed normally.
Displays information only for the paths accessing the specified host device. Specify the name of the logical device file (minus the slice number, in cUtXdY format) for the HDLM device to indicate the desired host device. The host-device-name string is case-sensitive. -stname Use this parameter to display the model ID of the storage system in the product ID element of the DskName field. When this parameter is omitted, the command displays the product ID or emulation type of the storage system instead.
To display path information (by selecting a display item) -path -item When you specify the -path parameter together with the -item parameter, the command only displays the items specified by the value of the -item parameter. If you specify no value for the -item parameter, only the PathID and the Status fields are displayed.
Selectable items Sub-parameters Physical-DskName vid Physical-iLU vid Physical-ChaPort vid Org-DskName ha Org-iLU ha # Because both PathID and Status are always displayed, you do not have to specify any parameters. In the sub-parameters following -path -item, you can list the paths (hdev) and sort the list (-srt). If you omit both parameters, the command displays information for all the paths in ascending order of AutoPATH_IDs.
• lu: LU number of the storage system • cp: Port number of the CHA When the -srt parameter is omitted, the path information is listed in order of ascending AutoPATH_IDs. When you use this parameter, the items used for the sorting keys (DskName, AutoPATH_ID, and the item specified by this parameter) are displayed by default. Therefore, it is not necessary to specify these items for the -item parameter. -t Omits the title for each information item.
shortened so that each instance of path information is displayed on a single line. For details on what is displayed for in each item, see Table 6-12 Displayed Path Information on page 6-47. When you use the -c parameter, the number of characters that can be displayed in the product ID element of the DskName field is limited to 10. When there are 11 or more characters in the product ID, the 8th and following characters are abbreviated to an ellipsis (...).
Displayed path information Table 6-12 Displayed Path Information on page 6-47 describes the displayed path information. The following explains the table headings: • No summary displayed: The user specifies only the -path parameter or the -path -item parameter. • Summary displayed: The user specifies the -path -c parameter. Table 6-12 Displayed Path Information Displayed Item No Summary summary displayed displayed Description Paths Total number of displayed paths, indicated by a decimal number.
Displayed Item No Summary summary displayed displayed Description IO-Count Total I/O count for all the displayed paths, indicated by a decimal number. The maximum value that can be displayed is 232 - 1 (4294967295). If the total I/O count reaches the maximum value, it will re-start from 0. IO-Errors Total I/O error count for all the displayed paths, indicated by a decimal number. The maximum value that can be displayed is 232 - 1 (4294967295).
Displayed Item No Summary summary displayed displayed Description This number combined with the storage system name (shown in DskName) identifies the LU that is accessed by a path. ChaPort#1 CP • For the Lightning 9900 Series, Lightning 9900V Series, Hitachi NSC55, Hitachi Universal Storage Platform 100, Hitachi Universal Storage Platform 600, Hitachi Universal Storage Platform 1100, or HUS VM indicated by a hexadecimal number.
Displayed Item No Summary summary displayed displayed Description When connecting to Lightning 9900 series, Lightning 9900V series, Hitachi USP series, Universal Storage Platform V/VM series, Virtual Storage Platform series, Hitachi AMS2000 series#2, Hitachi SMS series#2, HUS100 series#2, or HUS VM all paths are owner paths. IOCount#1 -- Total I/O count for the path, indicated by a decimal number. The maximum value that can be displayed is 232 1 (4294967295).
Displayed Item No Summary summary displayed displayed Description • * Indicates that an intermittent error occurred (automatic failbacks do not check the path) (the path status is Online, Online(E), Offline(E), or Offline(C)). HBAPortWW -N#1 A 16-digit hexadecimal number indicating the WWN information for an HBA connected to the storage system. This item is displayed only when you specify -iem together with the -hbaportwwn parameter.
The path information is displayed only when a value is specified for the path -item parameter. #2 This storage system applies when the dynamic I/O path control function is disabled. Table 6-13 Elements of a Path Name Element Corresponding information Solaris representation Host port number Host port number A unique number in a host: HDLM assigns this unique number to each HBA port. When two HBAs that have two ports are installed on the host, a value from 0000 to 0003 is assigned to each port.
Product ID Model names of storage systems • Without the -stname parameter With the -stname parameter (Displays the following for the model ID) Hitachi Universal Storage Platform 1100 SVS SVS • Hitachi Universal Storage Platform V USP_V • Hitachi Universal Storage Platform VM • Hitachi Virtual Storage Platform VSP HUS VM HUS_VM • P9500 P9500 XP48 XP48 XP128 XP128 XP512 XP512 XP1024 XP1024 XP10000 XP10000 XP12000 XP12000 XP20000 XP20000 XP24000 XP24000 # When a summary is d
displayed. The sorting key is iLU and its configuration information is displayed for each LU. By using the sub-parameter, -hdev or -pathid, you can filter the LU information to be displayed. If you do not specify the -hdev or -pathid parameter, the information about all LUs recognized by HDLM is displayed. For details on the content of each displayed item, see Table 6-17 Displayed LU Information on page 6-58.
To display LU information (by adding items to be displayed) -lu -item The items specified with the -item parameter are added to the items to be displayed by the -lu parameter and displayed. When the value of the -item parameter is omitted or all is specified, all the items, except for DPC, Physical-Product, Physical-SerialNumber, Physical-iLU, and Physical-ChaPort, that can be displayed are displayed.
For details on the contents of each displayed item, see Table 6-17 Displayed LU Information on page 6-58. The following shows the sub-parameters: -hdev host-device-name Filters the information only for the LUs corresponding to the specified host device. Specify the name of the logical device file (minus the slice number, in cUtXdY format) for the HDLM device to indicate the desired host device. The host-device-name string is case-sensitive.
Example 2 When using Hitachi AMS2000 series, Hitachi SMS series, or HUS100 series and displaying LU information with DPC added to the display items: # /opt/DynamicLinkManager/bin/dlnkmgr view -lu -item dpc Product : HUS100 SerialNumber : 9203008 LUs : 3 Dynamic I/O Path Control : on* iLU HDevName DPC PathID Status 000006 c6t5d0 on 000000 Online 000003 Online 000007 c6t5d1 off 000001 Online 000004 Online 000008 c6t5d2 on 000002 Online 000005 Online # To display a summary of LU information -lu -c When the -c
The items specified with the -item parameter are added to the items to be displayed by the -lu -c parameter and displayed. If no value is specified for the -item parameter, all the items that can be added are displayed. For the contents of each displayed item, see Table 6-17 Displayed LU Information on page 6-58. The following table lists the correspondence between the display items that can be added by using the -item parameter and the parameter values that can be specified after the -item parameter.
Displayed item No summary displayed Serial number Description Summary displayed S/N Serial number of the storage system Total number of LUs managed by HDLM among the LUs in the storage system LUs Dynamic I/O Path Control - The setting information about the dynamic I/O path control function is displayed for each storage system. on: The dynamic I/O path control function is enabled. off: The dynamic I/O path control function is disabled. -: The dynamic I/O path control function is not supported.
Displayed item No summary displayed Description Summary displayed number, and the last two characters are the internal LU number within the CU. SLPR#1 SLPR#2 The number of the SLPR to which an LU belongs, indicated by a number from 0 to 31. A hyphen (-) is displayed if the storage logical partition function for the storage system for the target LU is not supported. HDevName -- Host device name.
Displayed item No summary displayed ChaPort#1 Description Summary displayed -- Port number of the CHA, which identifies the CHA port that is mounted on the storage system. You can identify an actual CHA port by referencing this number from the storage system management program. CLPR#1 Status -- -- The number of the CLPR to which the CHA port belongs, indicated by a number from 0 to 31.
Displayed item No summary displayed IO-Errors#1 Description Summary displayed -- Total I/O error count for a path. The maximum value that can be displayed is 232 - 1 (4294967295). If the total I/O error count reaches the maximum value, it is reset, and the count is restarted from 0. To reset the IO-Errors value to 0, execute the dlnkmgr command's clear operation. Executing the clear operation also clears the number of I/O operations (IO-Count) to 0.
Displayed item No summary displayed Description Summary displayed CU number, and the last two characters are the internal LU number within the CU. • For Hitachi Virtual Storage Platform, indicated by a hexadecimal number. The first two characters of iLU are the number of the logical DKC (Disk Controller), the middle two numbers are the CU number, and the last two characters are the internal LU number within the CU. When a virtual ID is not used, a hyphen (-) is displayed.
To display HBA port information -hba Use this parameter to display information about the paths that pass through each HBA port. HDLM displays the unique HBA port ID assigned to each HBA port. This HBA port ID is used in monitor operations, and can also be used in Offline and Online operations. The following table lists and describes the displayed items. Table 6-18 Items Displayed as HBA Port Information Item Description HbaID HBA port ID Port.
-srt pb Sorts the HBA port information by the host port numbers (Port) and bus numbers (Bus). -c Does not display the SCSI device configuration information (Physical Device). -portwwn Displays port WWN information for HBAs connected to the storage system using a 16-digit hexadecimal number. -t Omits the title for each information item. To display CHA port information -cha Use this parameter to display information about the paths that pass through each CHA port.
Item Description which case you should check the status of individual paths and take appropriate action for any paths that have an error status. Example The following shows the subsequent sub-parameters: -srt cp Sorts the CHA port information by the storage system model IDs (Product), storage system serial numbers (S/N), and CHA port numbers (ChaPort). -t Omits the title for each information item.
Item Description U: Controller number reserved by HDLM. X: Target ID or WWN (World Wide Name) of the sd or ssd device for the HDLM device Y: LUN of the sd or ssd device for the HDLM device Device Type of SCSI device (sd or ssd) and corresponding instance number LDEV The model ID, serial number, and iLU number for the storage system, separated by periods. Information about LDEV can identify an actual LU that is managed by HDLM.
{pn | lu | cp}] [-t] dlnkmgr view -path -c [-stname] [-srt {lu | cp}] [-t] dlnkmgr view -lu [ -hdev HostDeviceName | -pathid AutoPATH_ID ] [-t] dlnkmgr view -lu -item [ [slpr] [pn] [cp] [clpr] [type] [ic] [ie] [dnu] [iep] [pd] [dpc] [vid] | all ] [ -hdev HostDeviceName | -pathid AutoPATH_ID ] [-t] dlnkmgr view -lu -c [-t] dlnkmgr view -lu -c -item [slpr] [-t] dlnkmgr view -drv [-t] dlnkmgr view -hba [-srt pb] [-c | -portwwn] [-t] dlnkmgr view -cha [-srt cp] [-t] KAPL01001-I The HDLM command completed normal
To display the monitor operation format /opt/DynamicLinkManager/bin/dlnkmgr monitor -help Parameters Note If the number of paths to be displayed by the monitor operation is 500 or more, the execution of the monitor operation might be delayed as follows: ¢ The time displayed in StartTime might be later than the time the monitor operation is actually executed. When the number of paths is approximately 2000, the time delay will be about seven seconds, and when it is 4000, 20 seconds.
Specifies the HBA port ID of the HBA port for which I/O information is to be displayed. The subsequent parameters specify the display interval and count. The format for the subsequent parameters are as follows: -intvl interval Specifies the interval at which I/O information is to be displayed, in the range from 1 to 60 (seconds). When this parameter is omitted, the default (3 seconds) is applied. -count count Specifies the number of times I/O information is to be displayed, in the range from 1 to 999.
00:01:30 557 0 KAPL01001-I The HDLM command completed normally. Operation name = monitor, completion time = yyyy/mm/dd hh:mm:ss To display I/O information for each CHA port -chaid CHA-port-ID Specifies the CHA port ID of the CHA port for which I/O information is to be displayed. The subsequent parameters specify the display interval and count.
Example # /opt/DynamicLinkManager/bin/dlnkmgr monitor -chaid 1 -intvl 30 -count 3 ChaID : 00001 Product : USP SerialNumber : 1100 ChaPort : 1R StartTime : yyyy/mm/dd hh:mm:ss ElapsedTime IO-Count IO-Errors 00:00:30 1154 0 00:01:00 205 22 00:01:30 54 10 KAPL01001-I The HDLM command completed normally. Operation name = monitor, completion time = yyyy/mm/dd hh:mm:ss To display monitor operation format -help Displays the monitor operation format.
To Display the Format of the add Operation /opt/DynamicLinkManager/bin/dlnkmgr add -help Parameters To Add a Path Dynamically -path Indicates that the target of the operation is a path managed by HDLM. Example # /opt/DynamicLinkManager/bin/dlnkmgr add -path KAPL01161-I This operation will change the path configuration. Do you want to continue? [y/n]:y KAPL01162-I A path was added. (path ID = 00010, storage = HITACHI.HUS100.9100163, iLU = 0030) : KAPL01159-I Paths were added.
¢ If you do not specify a disk label, the command might take time to complete the processing when you execute the command. To Display the Format of the add Operation -help Use this parameter to display the add operation format. Example # /opt/DynamicLinkManager/bin/dlnkmgr add -help add: Format dlnkmgr add -path [-s] KAPL01001-I The HDLM command completed normally.
7 Utility Reference This chapter explains the utilities used by HDLM.
Overview of the Utilities HDLM provides the following utilities: • The DLMgetras utility for collecting HDLM error information When an error occurs, this utility collects the files that contain information to be submitted to your HDLM vendor or maintenance company. For details on the DLMgetras utility, see The DLMgetras Utility for Collecting HDLM Error Information on page 7-3.
dlmvxexclude Utility for Assisting Creation of the VxVM Configuration File on page 7-30. • The installhdlm utility for installing HDLM This utility performs a new installation, upgrade installation, or reinstallation of HDLM in unattended installation mode. For details on this utility, see The installhdlm Utility for Installing HDLM on page 7-32.
Format /opt/DynamicLinkManager/bin/DLMgetras {directory-to-which-collectedinformation-is-output [-f name-of-a-file-for-defining-the-information-to-be-collected]|-h} You can also use lower-case characters (dlmgetras) as follows: /opt/DynamicLinkManager/bin/dlmgetras {directory-to-which-collectedinformation-is-output [-f name-of-a-file-for-defining-the-information-to-be-collected]|-h} Parameters directory-to-which-collected-information-is-output Specify the output directory for the information that is to be
¢ If a hash mark (#) is encountered anywhere other than at the beginning of a line, it is assumed to be part of the path name. ¢ Only one file or directory can be specified per line. ¢ The root directory (/) cannot be specified. ¢ ¢ When a directory is specified, the DLMgetras utility collects all the files in that directory, including files contained in the directory's subdirectories.
• Name-of-the-directory-containing-a-system-crash-dump-file/name-ofthe-system-crash-dump-file This is a symbolic link file to a system crash dump file. Set all files in the directory specified by the DUMPADM_SAVDIR parameter that is in the /etc/dumpadm.conf file. • hbsa.tar.Z This file contains compressed error information of the Hitachi Command Suite products other than HDLM. This file is output only when using the Hitachi Command Suite product that is the target for collecting error information.
Output directory#1 Files Explanation managing the HDLM configuration execution hdlmtr[1-64].log Trace file installhdlm.log Execution log file of the installhdlm utility for installing HDLM var/opt/ DynamicLinkManager/log/ mmap hdlmtr.mm Trace management file opt/DynamicLinkManager/ config dlmfdrv.unconf HDLM driver unconfiguration definition file dlmfdrv.unconf.bak Previous generation of the HDLM driver unconfiguration definition file. dlmfdrv.unconf.bak is created if the dlmfdrv.
Output directory#1 integrated-trace-file-outputdirectory-specified-in-theHitachi-NetworkObjectplaza-Trace-Libraryutility Files integrated-trace-fileprefix-specified-in-theHitachi-NetworkObjectplaza-TraceLibraryutility2[1-16].log#4 Explanation Integrated trace file (Default: /var/opt/ hitachi/HNTRLib2/ spool/hntr2[1-16].log) (Default: var/opt/hitachi/ HNTRLib2/spool) (default: hntr) kernel/drv dlmadrv.conf HDLM alert driver configuration definition file dlmfdrv.
Output directory#1 Files Explanation jnic146x.conf, jfca.conf, lpfc.conf, lpfs.conf, qla2200.conf, qla2300.conf, or qlc.conf etc md.conf Upper limit information on metadevice, metaset scsi_vhci.conf Sun StorageTek Traffic Manager Software setup and configuration definition file sd.conf sd (SCSI driver) configuration definition file ssd.conf Ssd (SCSI driver) configuration definition file coreadm.conf Core file configuration definition file devlink.
Output directory#1 Files driver_aliases driver_classes inittab Explanation Solaris settings and configuration information files minor_perm name_to_major path_to_inst release etc/cfg/fp fabric_WWN_map etc/default init etc/dfs dfstab Command settings file for execution at system boot time fstypes Settings file for default file system type (NFS), and other file system types main.cf VCS settings file main.cf.previous Previous generation file of main.
Output directory#1 Files Explanation dlmmgr-hba.txt Output results of dlnkmgr view -hba -portwwn dlmmgr-lu.txt HDLM LU information dlmmgr-lu-all.txt HDLM LU information (including the number of times an intermittent error occurred) dlmmgr-path.txt HDLM path information dlmmgr-path-iem.txt HDLM path information (including the number of times an intermittent error occurred) dlmmgr-sys.txt HDLM system settings dlmmgr-drv.
Output directory#1 7-12 Files Explanation prtconf-pv.txt Hardware configuration information of a host (output results of the prtconf -pv command) prtconf-vP.txt List of driver information prtconfMemory.txt Physical memory size prtvtoc.txt Partition configuration information of all disks (output results of the prtvtoc command) ps-ef.txt Process information psrinfo-v.txt CPU information showrev-p.txt List of installed patches svcs-a.
Output directory#1 Files Explanation results of the zpool status command) (Not applicable) VxVM #6 VxVM information collection directory VXVM_pkginfo.txt VxVM package information VXVM_vxdmp.txt /dev/vx/dmp directory list VXVM_vxrdmp.txt /dev/vx/rdmp directory list VXVM_vxdsk.txt /dev/vx/dsk directory list VXVM_vxrdsk.txt /dev/vx/rdsk directory list VXVM_vxdisklist.txt VxVM disk list VXVM_vxdisklist-s.txt Output results of vxdisk -s list VXVM_vxdiskpath.
Output directory#1 var/vxvm#6 Files vxconfigd.log.txt Explanation Log of error messages The output destination is either /var/vxvm/ vxconfigd.log or /var/ adm/syslog/syslog.log, depending on the debug option setting for the user. var/adm/syslog syslog.log.txt Log of error messages The output destination is either /var/vxvm/ vxconfigd.log or /var/ adm/syslog/syslog.log, depending on the debug option setting for the user. (Not applicable) VCS#6 VCS information collection directory IOF_gabconfig-a.
Output directory#1 etc/lvm#6 Files SDS_metastat.txt Metadevice information SDS_metaset.txt Diskset information SDS_metastat-s.txt metastat -s disksetname information md.tab SDS or SVM input parameter file md.cf SDS or SVM configuration backup file mddb.cf Information on SDS or SVM status database copy storage locations mdlogd.cf Log collected from SDS or SVM (Not applicable) SVM#6 etc/opt/SUNWmd#6 SC3.0#6 Explanation SVM information collection directory SVM_pkginfo.
Output directory#1 Files Explanation SC3.0_dev_did_rdsk.txt List of directories and files in the /dev/did/rdsk directory All directories and files in the /etc/cluster directory etc/cluster#6 Cluster configuration information for Sun Cluster #1 An output directory is created in the specified directory when you extract the getras.tar.Z file. #2 The value N depends on the setting in the dlmwebagent.properties file. The default value is 2.
Applicable operation • Deleting an existing LU (in Solaris 10 and Solaris 11) Inapplicable operation • Changing the settings of an existing LU (changing an LUN mapping or storage port) • Deleting a path • Removing an existing LU from HDLM management # Solaris 8 and Solaris 9 are not supported when a dynamic path is added because a new HBA was added. HDLM can use an LU, once it is recognized by the HDLM driver, without restarting the host.
Examples Example 1: The -a option is specified: Information on the newly added LUs is applied to the HDLM driver based on the contents of the HDLM driver configuration definition file (/kernel/ drv/dlmfdrv.conf). In Solaris 10 or Solaris 11, information on the LUs that were disconnected from the host is deleted from the HDLM driver. In Solaris 8 or Solaris 9, information on LUs that are already managed by HDLM is not updated.
¢ ¢ ¢ ¢ ¢ ¢ ¢ ¢ To remove the LU managed by HDLM from HDLM management, you need to restart the host specifying the reconfiguration option. Do not use the dlmcfgmgr utility. If you want to specify that some of the newly added LUs are not to be managed by HDLM, you need to remove these LUs from HDLM management before executing the dlmcfgmgr utility. If the newly added LUs have been removed from HDLM management, reconfiguration of these LUs is not performed for the HDLM driver.
The dlminstcomp HDLM Component Installation Utility This utility complements HDLM installation. Execute this utility when both of the following conditions are satisfied: • Linkage with Global Link Manager is used • When HDLM is installed, the KAPL09241-W message is output# # Execute this utility after installing the JDK. For details about the JDK version to be installed, see JDK required for linking with other Hitachi Command Suite products in Hosts and OSs Supported by HDLM on page 3-3.
file, results in the creation of a list that shows the correspondences of the logical device file names of HDLM devices. # /opt/DynamicLinkManager/bin/dlmlisthdev /var/tmp/ old_view_path_list.txt When upgrading HDLM 04-00 to a newer version of HDLM in a Solaris 8 environment The upgrade procedure is as follows: 1. Save the logical device file names of the HDLM devices used in HDLM 04-00.
Table 7-3 Information Displayed by the dlmlisthdev Utility for Assisting HDLM Transitions Item 6. Explanation OldHDevName The logical device file name (HDevName) of an HDLM device in HDLM 04-00 is displayed. If a hyphen (-) is displayed, it means that the corresponding HDev did not exist before the newer version of HDLM was installed. NewHDevName The logical device file name (HDevName) of an HDLM device of the newly installed version of HDLM is displayed.
Format /opt/DynamicLinkManager/bin/dlmsetboot -d post-migration-boot-diskdevice-name Parameters -d post-migration-boot-disk-device-name Specifies the HDLM device name of the post-migration boot disk. Notes • You cannot concurrently execute more than one dlmsetboot utility on the same host. In addition, you cannot execute this utility with another HDLM utility at the same time.
driver unconfiguration definition file (/opt/DynamicLinkManager/config/ dlmfdrv.unconf) is created. All HDLM-managed devices are defined in the HDLM driver unconfiguration definition file (/opt/DynamicLinkManager/config/dlmfdrv.unconf). The devices in the file are managed in units of LUs and begin with a hash mark (#) as comment lines.
definition file will be created in the /kernel/drv directory. Create the output directory before you execute the utility. -c Specify this parameter to check the integrity of the storage-systemmigration definition file used when the -t parameter is specified. If the -c parameter is specified, the utility reads the file, compares the existing dlmfdrv.conf and the list of physical paths (storage system information) in the destination system against the information that was read from the file.
definition file (/kernel/drv/dlmfdrv.conf) is not updated. If the HDLM driver unconfiguration definition file (/opt/DynamicLinkManager/config/ dlmfdrv.unconf) contains an invalid entry or does not exist, an error occurs. -t Specify this parameter to inherit the device-name properties (logical device names) in the existing dlmfdrv.conf as the device-name properties (logical device names) of the logical units in the destination storage system.
The -r parameter is specified: A new configuration definition file is created in the /kernel/drv directory. Existing information is not inherited. # /opt/DynamicLinkManager/bin/dlmsetconf -r Example 4 The -u parameter is specified: Defines a non- HDLM-managed LU based on the contents of the definitions in the HDLM driver unconfiguration definition file (/opt/ DynamicLinkManager/config/dlmfdrv.
7-28 • If the dlmsetconf utility finds an added path, in the HDLM driver configuration definition file (/kernel/drv/dlmfdrv.conf) the utility adds an entry (that is, a definition) to the definition of the LU for which the path has been added. If any path has been deleted, the entry (definition) is deleted from the definition of the affected LU. Whether a path is added or deleted, existing instance numbers are inherited unless a new HDLM driver configuration definition file (/kernel/drv/dlmfdrv.
• If the below message appears after you execute the dlmsetconf utility, it means that the device ID could not be acquired from the SCSI driver. If this happens, restart the host and then re-execute the dlmsetconf utility. KAPL10237-E The property information (devid) on a device (aa...aa) could not be collected. aa...aa: sd or ssd device name If the device ID cannot be acquired from the SCSI driver, refer to the bug report for Bug ID 4897065 provided by My Oracle Support (https:// support.oracle.
The dlmvxexclude Utility for Assisting Creation of the VxVM Configuration File This utility outputs the file (vxvm_exclude.txt) having the same format as the VxVM management target exclusion file (/etc/vx/vxvm.exclude)that matches the definition in the HDLM driver configuration definition file (/ kernel/drv/dlmfdrv.conf) to the directory specified by the -d parameter.
Display example Example The entries for the following devices are output: ¢ sd or ssd devices for HDLM-managed devices HDLM devices that are not actually created Only one HDLM device is created for one HDLM-managed device (one slice). However, VxVM assumes that the number of HDLM devices created for one HDLM-managed device is the same as the number of physical paths. Thus, executing the dlmvxexclude utility outputs entries for HDLM devices that have not actually been created.
The installhdlm Utility for Installing HDLM You can use the installhdlm utility to perform a new installation, upgrade installation, or re-installation of HDLM in unattended installation mode. You must predefine how the utility is to respond during installation in the installation-information settings file. Solaris 11 has functionality to create a new boot environment.HDLM, however, can be installed only in the current boot environment.
Table 7-4 Keys that Can Be Defined in the [INSTALLATION_SETTINGS] Section Required or optional? Key name#1 Explanation#2 installfile_location Specify the absolute path name of the directory on which the DVD-ROM is mounted. Maximu For m length upgrade For new (characte installatio installati rs) n or reon installatio n Optional Optional 400 Optional Optional 400 The default is /cdrom/ cdrom0. Specify the absolute path name of the directory to which log data will be output.
Required or optional? Key name#1 Explanation#2 Maximu For m length upgrade For new (characte installatio installati rs) n or reon installatio n HDLM driver is not configured regardless of the specified value. restart Specify whether to restart the host after installation by specifying either of the following values: Optional Optional 1 y: Restart the system. n: Do not restart the system (default). If you use Sun Cluster, specify n.
Table 7-5 Keys that Can Be Defined in the [ENVIRONMENT_SETTINGS] Section Required or optional? Key name#1 load_balance Explanation#2 Required Maximum for length Required upgrade (characte for new installati rs) installatio on or ren installati on Specify whether to enable Optional load balancing by specifying either of the following values: Optional 3 Optional 5 Optional Optional 6 Optional Optional 6 on: Enable load balancing (default). off: Do not enable load balancing.
Required or optional? Key name#1 Explanation#2 Required Maximum for length Required upgrade (characte for new installati rs) installatio on or ren installati on You can specify a value from 0 to 999999. The default is 100. Optional Optional 1 Optional Optional 7 Specify the maximum number Optional of error log files (dlmmgr[1-16].log). Optional 2 Optional Optional 1 Optional Optional 5 trace_file_number Specify the maximum number Optional of trace files (hdlmtr[1-64].log).
Required or optional? Key name#1 Explanation#2 Required Maximum for length Required upgrade (characte for new installati rs) installatio on or ren installati on on: Enable path health checking (default). off: Disable path health checking. path_health_check Specify the interval at which _interval the path health check will be performed in minutes.
Required or optional? Key name#1 Explanation#2 Required Maximum for length Required upgrade (characte for new installati rs) installatio on or ren installati on HDLM to assume an intermittent error. You can specify a value from 1 to 99. The default is 3. dynamic_io_path_c Specify whether to enable or ontrol disable the dynamic I/O path control function by using the values below. Note that, if this function is set, the setting for each storage system or LU is cleared.
For details on the settings functionality, see set (Sets Up the Operating Environment) on page 6-17.
About the Log File During an unattended installation, information about installation progress is output to the log file installhdlm.log. The following describes the installhdlm.log file: • The installhdlm.log file is created in the directory specified by the logdir key in the installation-information settings file. Note that the directory must not be shared with hosts on which unattended installation is performed. If the logdir key does not have a value, the file is created in the /var/tmp directory.
Log file The installux.sh utility outputs execution information to the log file /var/ tmp/hdlm_installux_sh.log. If the hdlm_installux_sh.log log file already exists, the utility appends execution information onto the end of the log file. If the size of the log file is equal to or greater than 1 MB when the utility is executed, the log file is renamed to hdlm_installux_sh2.log, and a new log file is created and named hdlm_installux_sh.log, to which the execution information is output.
7-42 Utility Reference Hitachi Dynamic Link Manager User Guide for Solaris®
8 Messages This chapter describes the format and meaning of the message IDs, and also the terms used in the messages and message explanations. For details on the meaning of the return codes output by HDLM when it receives a request from Global Link Manager and measures to take for them, see Return Codes for Hitachi Command Suite Common Agent Component on page 8-128.
□ Return Codes for Hitachi Command Suite Common Agent Component 8-2 Messages Hitachi Dynamic Link Manager User Guide for Solaris®
Before Viewing the List of Messages This section explains the following information that is needed to locate messages and understand the explanations in the sections from KAPL01001 to KAPL02000 on page 8-4KAPL01001 to KAPL02000. • Format and meaning of the message IDs • Terms used in the messages and message explanations This information is explained below. Format and Meaning of Message IDs Each message has a message ID. The following table shows the format and meaning of message IDs.
• HDLM management target KAPL01001 to KAPL02000 Message ID KAPL01001 -I Message Text The HDLM command completed normally. Operation name = aa...aa, completion time = bb...bb Explanation Details The HDLM command completed successfully. aa...aa: clear, help, monitor, offline, online, set, or view bb...bb: Year/month/day hour:minute:second Action None. KAPL01002 -I The HDLM command started. Operation name = aa...aa Details The HDLM command was executed. aa...
Message ID Message Text Explanation operation, see help (Displays the Operation Format) on page 6-4. KAPL01006 -W A necessary parameter is not specified. Operation name = aa...aa Details The specified operation does not contain the necessary parameter. aa...aa: clear, set, monitor, offline, or view Action Execute help operation-name of the HDLM command (dlnkmgr) to check the parameter. Specify the correct parameter, and then retry.
Message ID Message Text Explanation aa...aa: view Action Execute the view operation of the HDLM command (dlnkmgr) to check whether the HDLM manager has started. Start the HDLM manager if it has not started, and then retry the HDLM command. For details on the view operation, see view (Displays Information) on page 6-32. KAPL01013 -E An error occurred in internal Details processing of the HDLM command. An error unrelated to a user Operation name = aa...aa details = operation occurred during bb...
Message ID Message Text Explanation Execute the view operation of the HDLM command (dlnkmgr view -path) and check the value displayed in PathName. Use the two leftmost digits of PathName for the relevant HBA port, and then retry. For details on the view operation, seeview (Displays Information) on page 6-32. KAPL01016 -W The target CHA port was not found. Details Operation name = aa...aa The path ID indicated by -pathid and required by the -cha parameter is not an object of HDLM management. aa...
Message ID Message Text Explanation Execute the view operation of the HDLM command (dlnkmgr view -path) to check the value displayed in HDevName. Specify a host device for the value of HDevName, and then retry. For details on the view operation, seeview (Displays Information) on page 6-32. KAPL01019 -W The target path was not found. Operation name = aa...aa Details aa...aa: offline, online, monitor, or view ¢ offline/online operation The specified path does not exist.
Message ID KAPL01021 -E Message Text Cannot execute the HDLM command due to insufficient memory. Explanation Details Memory required for HDLM command processing could not be allocated. Action Terminate unnecessary applications to increase the amount of free memory, and then retry. KAPL01023 -W The last Online path for the device cannot be placed Offline(C). Details The path specified for the offline operation cannot be set to Offline(C) because it is the last, available path to the LU from the host.
Message ID Message Text continue the processing of the online operation? [y/n]: Explanation To ignore this path and perform online processing for the next path, enter y. To cancel processing, enter n. aa...aa: Path ID (decimal (base-10) number) Action If you want to continue processing of the online operation of the HDLM command for other paths, enter y. If you want to terminate processing, enter n. For details on the online operation, see online (Places Paths Online) on page 6-11.
Message ID Message Text Explanation Execute help operation-name of the HDLM command (dlnkmgr) to check the parameter value, and then retry. For details on the help operation, see help (Displays the Operation Format) on page 6-4). KAPL01048 -W Help information cannot be found. Operation name = aa...aa. Details The specified operation is not an operation of the HDLM command. aa...aa: Specified operation name Action Use the help operation of the HDLM command (dlnkmgr) to check the operation name.
Message ID KAPL01051 -I Message Text Because no path has been selected among the currently displayed paths, the paths in the Offline(C), Offline(E), and Online(E) statuses will be changed to the Online status. Is this OK? [y/n]: Explanation Details All the paths will be placed Online because the path selection parameter is not specified for the online operation. To place all the paths Online, enter y. To cancel the operation, enter n. Action If you want to execute online processing, enter y.
Message ID Message Text a path that cannot be placed in the Offline(C) status was detected. PathID = aa...aa Would you like to continue the processing of the offline operation? [y/n]: Explanation A path that cannot be placed set to Offline(C) was detected during multipath offline processing. To ignore this path and perform offline processing for the next path, enter y. To cancel offline processing, enter n. aa...
Message ID Message Text the Online status. Is this OK? [y/ n]: Explanation or -cha parameter was specified. If you want to continue, enter y. To cancel the operation, enter n. aa...aa: CHA port, HBA Action If you want to execute online processing for the paths that meet the specified requirements, enter y. If you want to terminate processing, enter n. KAPL01058 -W The specified parameter value is not needed. Operation name = aa...aa, parameter = bb...bb, parameter value = cc...
Message ID Message Text Explanation aa...aa: online, offline, set, or clear Action None. KAPL01061 -I aa...aa path(s) were successfully Details placed bb...bb; cc...cc path(s) were This message indicates the not. Operation name = dd...dd number of the paths processed in an online/offline operation. aa...aa: Number of paths where the online/offline operation was successful (decimal (base-10) number) bb...bb: Online, Online(S), Online(D) or Offline(C) cc...
Message ID Message Text Explanation To change the status of a path from Online(S) or Online(D) to Online, re-execute the HDLM command using the -hapath parameter. KAPL01068 -I Enter a license key: Details The license key will now be renewed. Enter a license key. Action None. KAPL01069 -W The entered license key is invalid. Details The entered license key is invalid. Action Enter a valid license key. KAPL01070 -E The entered license key is invalid. Renewal of the license key will now stop.
Message ID KAPL01074 -E Message Text The emergency license expired. Explanation Details The emergency license has expired. Register a permanent license. Action Register a permanent license. KAPL01075 -E A fatal error occurred in HDLM. The Details system environment is invalid. The license information file is missing. Action Re-install HDLM. KAPL01076 -I The permanent license has been installed. Details You need not install a license because a permanent license has already been installed. Action None.
Message ID Message Text Explanation is-to-occur-for-intermittent-errormonitoring). KAPL01081 -E The license key file is invalid. File name = aa...aa Details The format of the license key file is invalid. aa...aa: /var/tmp/hdlm_license Action Save the correct license key file in the designated, and then reexecute. /var/tmp/hdlm_license KAPL01082 -E There is no installable license key in the license key file. File name = aa...aa Details There is no useable license key for HDLM in the license key file.
Message ID Message Text Explanation cc...cc: Parameter values that cannot be specified at the same time Action Execute help operation-name of the HDLM command (dlnkmgr) to check which parameter can be specified, and then retry. For details on the help operation, see help (Displays the Operation Format) on page 6-4. KAPL01089 -E One of the following was executed at the same time as an HDLM command set -lic operation: another set -lic operation, or an update of the license for an update installation.
Message ID Message Text Explanation bb...bb: Code showing reason for error Action Re-execute. If the error occurs again, confirm that the system memory is sufficient. KAPL01092 -E The Offline path cannot be placed Online because a Reservation Conflict occurred. path ID = aa...aa, Details = bb...bb Details An attempt to recover the path failed because the LU for the path to be put online has been reserved by another path. aa...aa: Path ID bb...
Message ID Message Text Explanation An attempt to recover the path failed because an error occurred while accessing the SCSI driver. aa...aa: Path ID bb...bb: Code showing reason for error Action Re-execute. If the same error occurs even after removing the reservation, execute the DLMgetras utility for collecting HDLM error information to collect the error information, and then contact your HDLM vendor or maintenance company if you have a maintenance contract.
Message ID Message Text Explanation Action If you want to execute the operation of the HDLM command, enter y. If you want to terminate processing, enter n. KAPL01100 -I aa...aa Details This message indicates the executed command line. aa...aa: Name of the executed command. Action None. KAPL01101 -W The target HBA port was not found. Details Operation name = aa...aa The HBA having the HBA_ID specified in the -hbaid parameter could not be found. aa...
Message ID Message Text Explanation aa...aa: CHA, HBA Action If you want to execute the offline processing for the paths which pass the specified target, enter y. If you want to terminate the processing, enter n. KAPL01104 -I All the paths which pass the specified aa...aa port will be changed to the Online status. Is this OK? [y/n]: Details Multiple paths will be collectively placed Online because the -hbaid or -chaid parameter was specified. To collectively place multiple paths Online, enter y.
Message ID Message Text Explanation KAPL01114 -W The audit log configuration file does Details not exist. Restart the HDLM The audit log configuration file Manager, and execute the "dlnkmgr does not exist. view -sys -audlog" command and Action check the setting. Restart the HDLM Manager, and execute the "dlnkmgr view -sys -audlog" command, and then specify the desired setting by using the "dlnkmgr set audlog" command or the "dlnkmgr set -audfac" command as necessary.
Message ID Message Text Explanation Contact your HDLM vendor or the maintenance company if there is a maintenance contract for HDLM. KAPL01119 -W An error occurred during processing Details to output the audit log An internal parameter error when configuration file. the audit-log data was output. Action Contact your HDLM vendor or the maintenance company if there is a maintenance contract for HDLM. KAPL01120 -W A storage system model ID could not be displayed. Details = aa...aa, bb...
Message ID Message Text Explanation dd...dd: Online(S), Online(D) or Online(S)/Online(D) ee...ee: The number of paths which failed to change to either the Online, Online(S) or Online(D) status ff...ff: online Action None. KAPL01134 -I The target paths are already Online or Online(S). Details The specified paths are already in the Online or Online(S) status as a result of an online operation. Action Check path status by using the view operation..
Message ID Message Text Explanation Use the view operation of the HDLM command (dlnkmgr) to check the setting for storage and LUs. For details on the view operation, see view (Displays Information) on page 6-32. KAPL01157 -I The dynamic I/O path control function was set to aa...aa for the system. Details aa...aa: "on" or "off" Action Use the view operation of the HDLM command (dlnkmgr) to check the setting for the host, storage, and LUs.
Message ID KAPL01161 -I Message Text This operation will change the path configuration. Do you want to continue? [y/n]: Explanation Details This message confirms whether to perform a path configuration change by using an add operation. Action Enter y to change the path configuration, or enter n to cancel the operation. KAPL01162 -I A path was added. (path ID = aa...aa, storage = bb...bb, iLU = cc...cc) Details This message displays information about a path added as the result of an add operation. aa...
Message ID Message Text Explanation path can be used for the system, enter y. To cancel the operation, enter n. KAPL01167 -I All paths will be set to Online or Online(D). Is this OK? [y/n]: Details All paths will be set to Online or Online(D) because no path is specified. To continue, enter y. To cancel the operation, enter n. Action To set all paths to Online or Online(D), enter y. To cancel the operation, enter n.
Message ID Message Text Explanation To continue, enter y. To cancel the operation, enter n. Action To set to online the all Online(S) or Online(D) paths that are connected to the specified LU with path ID, enter y. To cancel the operation, enter n. KAPL01171 -I The target paths are already aa...aa or bb...bb. Details The specified paths are already in the Online, Online(S), or Online(D) status as a result of an online operation. aa...aa: Online or Online(S) bb...
Message ID KAPL03003 -E Message Text HDLM API Error information aa...aa Explanation Details This information is required for resolving problems. aa...aa: API trace error information Action Execute the DLMgetras utility for collecting HDLM error information, and then contact your HDLM vendor or the maintenance company if there is a maintenance contract for HDLM. For details on the DLMgetras utility, see The DLMgetras Utility for Collecting HDLM Error Information on page 7-3.
Message ID Message Text Explanation HDLM Error Information on page 7-3. KAPL03007 -E An error occurred during communication with the HDLM manager. (aa...aa) Details This information is required for resolving problems. aa...aa: API trace error information Action Execute the DLMgetras utility for collecting HDLM error information, and then contact your HDLM vendor or the maintenance company if there is a maintenance contract for HDLM.
Message ID Message Text Explanation HDLM Error Information on page 7-3. KAPL04001 to KAPL05000 Message ID Message Text Explanation KAPL04001 HDLM manager started. -I Action KAPL04002 Could not start the HDLM manager. -E Details None. The HDLM manager failed to start because the current environment is unsuitable for the HDLM manager to run in.
Message ID Message Text Explanation HDLM manager could not start normally because it was unable to open the option definition file. aa...aa: Option definition file name Action Check whether another program is currently using this file (for example, the file is being opened in a text editor), or whether the file has been inadvertently deleted. KAPL04009 The option definition is invalid.
Message ID Message Text KAPL04011 Could not output the error log file. -E Explanation Details The log information could not be output to the error log file /var/ opt/DynamicLinkManager/log/ dlmmgr[1-16].log. Action Make sure that the disk has enough free capacity. KAPL04012 Could not create a communication -E pipe. RC = aa...aa Details HDLM manager could not start normally because it was unable to create a pipe file, which is used in communication with HDLM commands. aa...
Message ID Message Text Explanation Data could not be written to the pipe file while communicating with an HDLM command. aa...aa: OS error code (decimal (base-10) number) Action Execute the DLMgetras utility for collecting HDLM error information, and then contact your HDLM vendor or the maintenance company if there is a maintenance contract for HDLM. For details on the DLMgetras utility, see The DLMgetras Utility for Collecting HDLM Error Information on page 7-3.
Message ID Message Text Explanation Execute the DLMgetras utility for collecting HDLM error information, and then contact your HDLM vendor or the maintenance company if there is a maintenance contract for HDLM. For details on the DLMgetras utility, see The DLMgetras Utility for Collecting HDLM Error Information on page 7-3. KAPL04023 HDLM manager error information -E aa...aa Details This information is required for resolving problems. aa...
Message ID Message Text Explanation To increase the amount of free memory, terminate unnecessary applications or restart the host. KAPL04026 The temporary license is valid. The -I license expires in aa...aa days on (bb...bb). Details aa...aa: Number of days remaining bb...bb: Expiration date format: Year (4 digits)/month (01-12)/ day (01-31) Action Install a permanent license by the expiration day. KAPL04027 The emergency license is valid. The -I license expires in aa...aa days on (bb...bb).
Message ID Message Text Explanation As for the options other than the defaults, use the dlnkmgr set operation to set the options again.For details on the set operation, see set (Sets Up the Operating Environment) on page 6-17. KAPL04034 An attempt to create the option -E definition file has failed. Details An attempt to re-create an option definition file (/etc/opt/ DynamicLinkManager/ dlmmgr.xml) using the default values has failed.
Message ID Message Text Port Number= cc...cc, Community = dd...dd, Trap Data = ee...ee Explanation aa...aa: Trap ID bb...bb: Destination IP address of the trap cc...cc: Destination port number of the trap dd...dd: Community name given to the trap ee...ee: Transmission data Action None. KAPL04053 The option definition file was -W recovered from a work file. Action KAPL04054 The owner controller will now be -I revised. (number of LUs = aa...
KAPL05001 to KAPL06000 Message ID Message Text KAPL05003 The HDLM driver (filter component) -I was successfully attached to Disk (aa...aa), Partition (bb...bb). Explanation Details The path for the disk and partition was registered successfully in the core logic. aa...aa: minor number of the sd or ssd device (decimal number) bb...bb: slice number Action None. KAPL05008 Could not allocate memory. -E (aa...aa:bb...
Message ID Message Text Explanation Check whether the HDLM driver has started normally. If it has not started or contains an error, contact your HDLM vendor or the maintenance company, if there is a maintenance contract for HDLM, and report the error and detail code. KAPL05014 The device object (aa...aa) was -I registered as the path (bb...bb). Details The HDLM filter driver was successfully attached and registered as a path. aa...
Message ID Message Text Explanation aa...aa: IOCTL code (hexadecimal number) Action None. KAPL05021 Processing of IOCTL(aa...aa) -I completed normally. Details The processing for the requested IOCTL operation was successful. aa...aa (hexadecimal number): IOCTL code Action None. KAPL05023 Could not process the -E IOCTL(aa...aa). (bb...bb:cc...cc) Check the message of the HDLM command or the HDLM manager, and then take the appropriate action.
Message ID Message Text Explanation Action Check whether the path is correctly connected to the disk. KAPL05040 The disk geometry or partition -E information could not be set. IOCTL = aa...aa, ErrorCode = bb...bb, PathID = cc...cc After fixing the path, change its status to Online. Refer to the HDLM User's Guide for instructions how to change the path status to Online.
Message ID Message Text Explanation aa...aa: Minor number of the sd or ssd driver (hexadecimal number) bb...bb: IOCTL type (internal code) (hexadecimal number) cc...cc: Error number (decimal number) dd...dd: uscsi_status (hexadecimal number) Action None. KAPL05411 The instance number aa...aa for -W "dlmfdrv" already exists. The HDLM device could not be attached. Execute the DLMgetras utility to collect error information, and then contact your vendor or maintenance company.
Message ID Message Text Explanation KAPL05437 The contents of the configuration Details -E definition file aa...aa.conf are invalid. An entry in a configuration Check whether the path is correctly definition file (/kernel/drv/ connected to the disk. Execute the dlmfdrv.conf or /kernel/drv/ dlmsetconf command and reboot. dlmndrv.conf) of the filter Refer to the HDLM User's Guide for driver or nexus driver is invalid. instructions how to execute the aa...aa: Driver name dlmsetconf command.
Message ID Message Text Explanation Stop any applications that are using the HDLM devices, and then re-execute the dlmcfgmgr utility for managing the HDLM configuration. KAPL05514 The label might be invalid or there -W might be a problem with the slice to be mounted. Details The label might be invalid or there might be a problem with the slice to be mounted. Action If the mounted LU is UFS, check the mounted slice. In all other cases, the created label might be invalid. Check the label.
Message ID Message Text Explanation Action Check the memory capacity which can be used. KAPL05903 Failed to search aa...aa directory. -E Check the permission for the directory. Details An attempt by dlmsetdevname to search for a directory has failed. aa...aa: Directory name Action Check the permission for the directory. KAPL05904 Failed to set device name. Execute -E the DLMgetras utility to collect error information, and then contact your vendor or maintenance company.
KAPL06001 to KAPL07000 Message ID Message Text Explanation KAPL06003 Initialization of the HDLM alert driver Details -I (aa...aa) was successful. aa...aa: Address of the table for managing the alert driver (hexadecimal number) Action None. KAPL06004 Could not allocate memory. -E (aa...aa:bb...bb) Details An attempt to reserve memory to save alert information has failed. aa...aa: Program line (hexadecimal number) bb...
Message ID Message Text Explanation A request to the alert driver was made with an invalid IOCTL code. aa...aa: IOCTL code (hexadecimal number) Action None. KAPL06010 Could not process the -E IOCTL(aa...aa). (bb...bb:cc...cc) Details An IOCTL request was not processed normally. aa...aa: IOCTL code (hexadecimal number) bb...bb: IOCTL end code (hexadecimal number) cc...cc: 0 (fixed) Action Check the message of the HDLM command (dlnkmgr) or HDLM manager, and then take the appropriate action.
Message ID Message Text Explanation Review the actual memory size when another error does not occur. When the actual memory size is insufficient, increase the actual memory size. When the actual memory size is sufficient, execute the DLMgetras utility for collecting HDLM error information, and then contact your HDLM vendor or the maintenance company if there is a maintenance contract for HDLM.
KAPL07001 to KAPL08000 Message ID KAPL07819 -I Message Text Data for maintenance: aa...aa bb...bb cc...cc dd...dd. Explanation Details This message is generated by the core logic for maintenance. aa...aa: Detailed information 1 (decimal (base-10) number) bb...bb: Internal function number of the core logic (decimal (base-10) number) cc...cc: Detailed information 2 (decimal (base-10) number) dd...dd: Detailed information 3 (decimal (base-10) number) Action None.
KAPL08001 to KAPL09000 Message ID Message Text KAPL08019 The path (aa...aa) detected an error -E (bb...bb). (cc...cc) Explanation Details An error occurred in the path. The error is most likely due to a disconnected cable. aa...aa: Path identifier (hexadecimal number) bb...bb: Error code (hexadecimal number) ¢ When a path error is detected by a path health checking or the online operation: Displays 0x000F0000(Fixed). ¢ When a path error is detected through an I/O error: Displays the OS error code.
Message ID Message Text Explanation ee...ee: Target ID (same as PathName of view -path) (hexadecimal number) ff...ff: HLU number (same as PathName of view -path) (hexadecimal number) gg...gg: Dev number (same as DNum of view -path) (decimal (base-10) number) hh...hh: Host device name (same as HDevName of view path) Action There could be an error in the path. See What To Do for a Path Error on page 5-3 and restore the path displayed in the message to running status. KAPL08023 A path was recovered.
Message ID Message Text Explanation aa...aa: Path ID (same as PathID of view -path) (decimal (base-10) number) Action Errors are detected in all the paths connected to the LUs. See What To Do for a Path Error on page 5-3 to make the path shown in the error message or the paths connected to the target LU. KAPL08027 A path was excluded from the items -E subject to automatic failback. PathID = aa...
Message ID Message Text Explanation the "dlnkmgr online -dfha" command. KAPL09001 to KAPL10000 Message ID KAPL09002 -E Message Text The disk does not have sufficient free space. Explanation Details The file system of the installation does not have sufficient free capacity to install HDLM. Action Delete unnecessary files to increase the amount of free capacity, and then retry. KAPL09008 -W The license code is invalid. KAPL09009 -E The license code is invalid.
Message ID Message Text Explanation there is a maintenance contract of HDLM. KAPL09020 -E An attempt to remove Hitachi Network Objectplaza Trace Library 2 failed. Details An attempt to remove HNTRLib2 has failed. Action Manually remove Hitachi Network Objectplaza Trace Library 2. If the attempt to remove Hitachi Network Objectplaza Trace Library 2 fails again, contact your HDLM vendor or the maintenance company if there is a maintenance contract for HDLM.
Message ID Message Text Explanation An attempt to delete a file related to HDLM from the directory of Hitachi Command Suite products other than HDLM has failed. Action If the following files or directories are on the host, remove them manually. /opt/HDVM/agent/docroot/ webstart/HDLM.jnlp /opt/HDVM/agent/classes/com/ Hitachi/soft/HiCommand/DVM/ agent/module/ HDLMManager.
Message ID Message Text Explanation KAPL09033 -E HDLM supports only SPARC Action platform systems. Please install See Hosts and OSs Supported by HDLM on a supported architecture. HDLM on page 3-3 and install HDLM in a supported OS system. KAPL09076 -I The permanent license was installed. Details The permanent license was installed. Action None. KAPL09077 -I The temporary license was installed. The license expires on aa...aa. Details A temporary license was installed. aa...
Message ID Message Text KAPL09082 -W The temporary license expired. KAPL09083 -W The emergency license expired. KAPL09086 -W The entered license key is invalid. KAPL09087 -E The entered license key is invalid. Renewal of the license key will now stop. Obtain a valid license key, and then re-install HDLM. Explanation Action Enter a permanent license key. Action Install a permanent license. Action Enter a valid license key.
Message ID Message Text Explanation designated directory (/var/ tmp/hdlm_license), in a user-specified directory or with user-specified file name, and then re-execute the HDLM installation program. KAPL09090 -W This operation will now be continued without updating the license. Details This operation will continue without updating the license. Action Install a permanent license at a later time. KAPL09091 -E A fatal error occurred in HDLM. Details The system environment is invalid.
Message ID Message Text Explanation /var/tmp/hdlm_license or the file that the user specified. KAPL09114 -I There is no license key file. File name = aa...aa Details There is no license key file in the designated directory. aa...aa: /var/tmp/hdlm_license Action Enter a license key when prompted by a message to do so. Alternatively, cancel the installation, save the correct license key file in the designated directory, and then re-execute the HDLM installation program.
Message ID Message Text Explanation Do not perform the following operation: Execution of the HDLM command's set -lic operation simultaneously with an update of the license for an upgrade or reinstallation. KAPL09137 -W An internal error occurred when the system information was set. Please contact product support. Details During the installation of HDLM, an error considered not to be the result of a user operation occurred.
Message ID Message Text Explanation Contact your HDLM vendor or the maintenance company if there is a maintenance contract for HDLM. KAPL09148 -W Could not modify file. File name = aa...aa, Details = bb...bb. Please contact product support. Details An HDLM driver configuration definition file or HDLM driver unconfiguration definition file, aa...aa, could not be updated. aa...aa: One of the following file names: ¢ ¢ ¢ ¢ /kernel/drv/dlmfdrv.conf /opt/DynamicLinkManager/ config/dlmfdrv.
Message ID Message Text Explanation For details on the dlmsetconf utility, see The dlmsetconf Utility for Creating the HDLM Driver Configuration Definition File on page 7-23. c. Execute either of the following commands to restart the host: # reboot -- -r boot -r (at the ok prompt) KAPL09149 -E An attempt to update the file has failed. File name = aa...aa, Details = bb...bb. Please contact product support.
Message ID Message Text Explanation Contact your HDLM vendor or the maintenance company if there is a maintenance contract for HDLM. KAPL09176 -I KAPL09179 -I The installation completed successfully. Execute the dlmsetconf utility and, if you are using VxVM, execute the dlmvxexclude utility. After that, specifying the -r (reconfigure) option, restart the host machine. Data for maintenance: aa...aa bb...bb Action a. Create HDLM driver configuration definition file (/ kernel/drv/dlmfdrv.
Message ID Message Text Explanation Make sure that the parameters for installhdlm are appropriate, and then try again. KAPL09188 -W Too many parameters are specified. Details More than two parameters have been specified for the installhdlm utility for installing HDLM. Action Make sure that the parameters for installhdlm are appropriate, and then try again. KAPL09190 -W The installation information settings file is not specified.
Message ID Message Text Explanation Follow the message output at the same time. KAPL09203 -W There is no license key file. File name = aa...aa Details There is no license key file. aa...aa: Character string that the user entered Action Follow the message output at the same time. KAPL09204 -E The entered absolute path is invalid. The HDLM installation will now terminate. Action KAPL09205 -E There is no license key file. The HDLM installation will now terminate. File name = aa...
Message ID KAPL09210 -I Message Text aa...aa will now start. Explanation Details aa...aa has just started. aa...aa: installhdlm, pkgadd, dlnkmgr, or dlmsetconf Action None. KAPL09211 -I aa...aa completed successfully. Details aa...aa completed successfully. aa...aa: installhdlm, pkgadd, dlnkmgr, or dlmsetconf Action None. KAPL09212 -E aa...aa ended abnormally. Details aa...aa ended abnormally. aa...
Message ID Message Text Explanation The system environment is not valid as an environment for executing the installhdlm utility. aa...aa: Error number (decimal number) that specifies the executed processing Action If the error code is 4: The /var/tmp directory does not exist. Prepare the /var/ tmp directory, and then retry the operation. If the error code is 5: You do not have write permission for the /var/tmp directory. Make sure you have access permissions for the / var/tmp directory.
Message ID Message Text Explanation aa...aa: Error number that indicates the executed processing (decimal (base-10) number) bb...bb: Return value of the executed processing (decimal (base-10) number) Action Make sure that sufficient free capacity exists for the /var/tmp directory. If the capacity is insufficient, allocate the required amount of capacity, and then retry the operation.
Message ID Message Text Explanation If aa...aa is dlnkmgr or dlmsetconf: The correct location of dlnkmgr and dlmsetconf is / opt/DynamicLinkManager/ bin. If dlnkmgr or dlmsetconf was not in the correct location, or you did not have a required permission, re-execute installhdlm. If aa...aa is installhdlm_analysis: The correct location of installhdlm_analysis is the same directory as installhdlm.
Message ID Message Text Explanation aa...aa: Error number (decimal number) that specifies the executed processing bb...bb: Return value (decimal number) of the executed processing Action Specify the revised installation information settings file, and then retry the operation. KAPL09221 -W The definition of the installation information settings file includes an unusable character. Error Code = aa...aa,bb...bb, line = cc...cc Details A character that cannot be used in a non-comment line was used. aa...
Message ID Message Text Explanation bb...bb: Return value (decimal number) of the executed processing cc...cc: Line number (decimal number) of an installation information settings file. Action Specify the revised installation information settings file and then retry the operation. KAPL09229 -W The definition of the installation information settings file includes an invalid section name. Error Code = aa...aa,bb...bb, line = cc...cc Details A section that does not exist is included. aa...
Message ID Message Text Explanation bb...bb: Return value (decimal number) of the executed processing cc...cc: Line number (decimal number) of an installation information settings file. Action Specify the revised installation information settings file and then retry the operation. KAPL09232 -W The composition of the definition of the installation information settings file is invalid. Error Code = aa...aa,bb...bb, line = cc...cc Details A key, key value, or equal sign (=) is missing. aa...
Message ID Message Text Explanation aa...aa: Name of the folder or file that does not exist. Action Specify the revised installation information settings file and then retry the operation. KAPL09235 -E The log file cannot be output to its destination because the environment is invalid. Error Code = aa...aa Details The log file cannot be output in this environment, because certain environment settings are invalid. aa...aa: Error number (decimal number) that specifies the executed processing.
Message ID KAPL09237 -I Message Text A user operation ended installhdlm. Explanation Details The process has been terminated by an operation such as Ctrl+C. Action Check if the /var/opt/ DynamicLinkManager directory exists. If the directory exists, delete it by using the following command: # rm -rf /var/opt/ DynamicLinkManager However, if you interrupted an upgrade installation or reinstallation, do not delete this directory. Also, execute the following command to check the status.
Message ID Message Text Explanation requires no particular action, because HDLM operation is not affected when linkage with Global Link Manager is not used. If the error code is 3 or 100: Wait a while, and then execute the HDLM component installation utility (dlminstcomp). If the error code is 5: Install the prerequisite JDK, and then execute the HDLM component installation utility (dlminstcomp).
Message ID Message Text Explanation aa...aa: Error number (decimal number) that specifies the executed processing Action Contact your HDLM vendor or your maintenance company if you have a maintenance contract for HDLM. KAPL09248 -W A parameter is invalid. parameter = aa...aa Details An invalid parameter was specified. aa...
Message ID Message Text removehdlm from the installation media. Explanation Action Execute removehdlm from the DVD-ROM used for removal. removehdlm: DVD-ROM-mountpoint/HDLM_Solaris/IPS/ remove_scripts/removehdlm KAPL09290 -I A user operation ended removehdlm. Details The process has been terminated by an operation such as Ctrl+C. Action Re-execute removehdlm. KAPL09291 -E This version of the OS is not supported. Please remove HDLM on a supported OS. Details This version of the OS is not supported.
Message ID Message Text Explanation remove the installed HDLM or SP, and then perform a new installation. When installing a SP: Obtain, and then install an SP or corrected version that can be applied to the installed version of HDLM. KAPL09509 -E Service Pack aa...aa cannot be installed. The same version has already been installed. Details The version of the SP being installed is the same as the already installed SP. Installation of the SP has stopped. aa...
Message ID KAPL09603 -E Message Text The installation will now stop because a command or utility cannot be executed. Confirm that the command or utility (aa...aa) exists, and that you have the proper permission. Explanation Details aa...aa: Name of the command or utility that cannot be executed. Action When the output information is an HDLM utility: ¢ ¢ If an installation is performed from the DVD-ROM, confirm that the DVD-ROM is not damaged.
Message ID Message Text Explanation DLMgetras utility, see The DLMgetras Utility for Collecting HDLM Error Information on page 7-3. KAPL10002 Too many parameters have been -W specified. Details Four or more parameters have been specified. Action Check the parameters of the DLMgetras utility for collecting HDLM error information, and then retry. For details on the DLMgetras utility, see The DLMgetras Utility for Collecting HDLM Error Information on page 7-3.
Message ID Message Text Explanation Action Check the parameters of the DLMgetras utility for collecting HDLM error information, and then retry. For details on the DLMgetras utility, see The DLMgetras Utility for Collecting HDLM Error Information on page 7-3. KAPL10006 The file for defining the information Details -W to be collected does not exist, or The file for defining the cannot be read. Value = aa...
Message ID Message Text Explanation b. Check whether the specified directory name is correct. c. Check that the disk has sufficient free capacity. KAPL10009 The specified directory already Details -W exists. Do you want to overwrite it? The specified directory already [y/n]: exists. Enter y to overwrite it, or n to cancel. Action The specified directory already exists. Enter y to overwrite the existing file.
Message ID Message Text Explanation After the DLMgetras utility for collecting HDLM error information terminates, check the contents of the file for defining the information to be collected. This file is shown in the message. If the contents of the file are incorrect, correct them and then try to collect error information again. The DLMgetras utility will ignore the specified file or directory and continue processing.
Message ID Message Text Explanation bb...bb: Indicated contents in a line Action After the DLMgetras utility for collecting HDLM error information terminates, check the contents of the file for defining the information to be collected. This file is shown in the message. If the contents of the file are incorrect, correct them and then try to collect error information again. The DLMgetras will ignore the specified file and continue processing.
Message ID Message Text Explanation Action After the DLMgetras utility for collecting HDLM error information terminates, check whether the file for defining the information to be collected is a text file. The file is shown in the message. For details on the DLMgetras utility, see The DLMgetras Utility for Collecting HDLM Error Information on page 7-3. KAPL10016 The root directory has been -W specified in the first parameter.
Message ID Message Text Explanation The utility for collecting HDLM error information terminated before completion. If the directory is unnecessary, delete directory. For details on the DLMgetras utility, see The DLMgetras Utility for Collecting HDLM Error Information on page 7-3. KAPL10022 The utility for collecting HDLM error Details -I information completed normally. Error information has been collected. Action None.
Message ID Message Text KAPL10033 The file does not exist. Filename = -W aa...aa Explanation Details The file to collect does not exist. aa...aa: File to collect Action None. KAPL10034 The file could not be copied. -E Filename = aa...aa, Details = bb...bb Details Execution of the cq command failed. aa...aa: File name you tried to copy bb...bb: cq command output message Action An error occurred during copying of the file to be collected. The user environment might have been unstable.
Message ID Message Text Explanation maintenance company if you have a maintenance contract for HDLM. KAPL10049 Error information collection -I command = aa...aa, Return value = bb...bb, Execution time = cc...cc Details A command was executed to collect information targeted for collection. aa...aa: Executed command bb...bb: Return value of the executed command cc...cc: year (4 digits)/month/ date hour:minute:second Action None.
Message ID Message Text Explanation None. KAPL10220 An error occurred when collecting Details -E the HDLM support disk information. An attempt to acquire information about HDLM-managed devices has failed. Action Check the system configuration. KAPL10221 An error occurred when creating -E the configuration definition file (aa...aa). Details An attempt to create a configuration definition file has failed. aa...aa: Name of the configuration definition file Action Check the system configuration.
Message ID Message Text Explanation bb...bb: Name of the configuration definition file Action Check the existing bb...bb file. KAPL10228 The HDLM support disk information -W does not exist. Details HDLM-managed devices cannot be found. Action Check the system configuration. KAPL10229 Internal processing error. -E Details An internal processing error occurred. Action Internal processing error. Contact your HDLM vendor or the maintenance company if there is a maintenance contract for HDLM.
Message ID Message Text Explanation Remove the error by referencing Errno, and then retry. KAPL10235 Get symbolic link file(aa...aa) -W status failed. Function = lstat, Errno = bb...bb (cc...cc) Details An attempt to acquire information about the symbolic link file has failed. aa...aa: File name bb...bb: Errno (decimal number) cc...cc: Errno string Action Remove the error by referencing Errno, and then retry. KAPL10236 Could not allocate memory. Errno -E = aa...aa (bb...
Message ID Message Text KAPL10240 The directory specified by the -d -E option does not exist. Explanation Details The directory specified in the -d option was not found. Action Check whether the specified directory exists. KAPL10241 An error occurred when collecting -E the catalog message. Errno = aa...aa (bb...bb) Details A message could not be acquired from the catalog file. aa...aa: Errno (decimal number) bb...bb: Errno string Action Remove the error by referencing Errno, and then retry.
Message ID Message Text Explanation path information was redefined on the basis of that information. Action Remove the physical path that was shown in the above KAPL10243-W messages from the configuration definition file if that physical path is not needed. KAPL10251 An attempt to retrieve information Details -W for the directory aa...aa failed. An attempt to collect directory Function = readdir, errno = bb...bb information has failed. (cc...cc) aa...aa: Directory name bb...bb: Errno (decimal number) cc.
Message ID Message Text Explanation If you want to continue the dlmsetconf utility processing, enter y. If you want to stop, enter n. KAPL10255 The entered value is invalid. The -E operation stops. Details Processing will stop because an incorrect value was entered three or more times for the message confirming the execution of the dlmsetconf utility for creating the HDLM driver configuration definition file. Action When executing the utility, enter the correct value.
Message ID Message Text Explanation deleted, no action is required. To prevent this message from being output, restart the host. If the disk indicated in Device is the LU that was not dynamically deleted, make sure that the disk is properly connected to the server. If there are any errors, remove the error, and then re-execute the command. If the disk is connected properly, contact your HDLM vendor or the maintenance company if there is a maintenance contract for HDLM.
Message ID Message Text Explanation Remove from the dlnkmgr view -path command error, and execute the command again. KAPL10299 The input file is invalid. The input -E file is not the output result of the "dlnkmgr view -path" command. Details The content of a file specified in a parameter of the dlmlisthdev command differs from the output of the dlnkmgr view -path command. Action Check the input file, and then retry.
Message ID Message Text Explanation KAPL10329 A file could not be opened. File Details -E name = aa...aa, Errno = bb...bb. The file could not be opened. Refer to the Messages section of aa...aa: File name the HDLM User's Guide for instructions to correct this problem. bb...bb: Errno (decimal number) Action Confirm the existence of the file and access permission. If there is no problem, contact your HDLM vendor or maintenance company if you have a maintenance contract for HDLM. KAPL10334 Ioctl aa...
Message ID Message Text Explanation None. For details on the dlmcfgmgr utility, see The dlmcfgmgr Utility for Managing the HDLM Configuration on page 7-16. KAPL10371 The HDLM support disk information -W does not exist in the configuration file (dlmfdrv.conf). Check the system configuration, and then retry the command. Details The /kernel/drv/dlmfdrv.conf file does not contain the data of the HDLM-managed devices. Action Check the system configuration, and then retry.
Message ID Message Text Explanation for managing the HDLM configuration. Action Enter y or n. For details on the dlmcfgmgr utility, see The dlmcfgmgr Utility for Managing the HDLM Configuration on page 7-16. KAPL10376 The device (aa...aa) could not be -E opened. ErrorCode = bb...bb (cc...cc) Details An attempt to open the device managing the HDLM filter driver has failed. aa...aa: Name of the device managing the HDLM filter driver bb...bb: Errno (decimal number) cc...
Message ID Message Text Explanation the HDLM configuration is incorrect. Action Specify appropriate parameters, and then retry. For details on the dlmcfgmgr utility, see The dlmcfgmgr Utility for Managing the HDLM Configuration on page 7-16. KAPL10382 There is no instance of the HDLM Details -W nexus driver that manages the path This LU was not added because of the LU. This LU (LDEV = aa...aa) the HDLM nexus driver instance is not added. FilterPath = bb...bb. does not exist in the LU.
Message ID Message Text Explanation None. KAPL10386 The LU (LDEV = aa...aa) is already -I managed by HDLM. If the path configuration of the LU shown by the LDEV name has changed, reboot the host. Details This LU is already managed in dlmfdrv. aa...aa: LDEV name Action If the path configuration of the LU shown by the LDEV name has changed, reboot the host. If there is no change to the configuration, do nothing. KAPL10387 The path of the another LU (LDEV -W = aa...
Message ID Message Text Explanation Check the capacity of the root file system or the file system including the file. Check the memory capacity which can be used. KAPL10391 The output file (aa...aa) could not -E be opened. ErrorCode = bb...bb (cc...cc). Resolve the error by referencing ErrorCode. Details An attempt to open the output file has failed. aa...aa: Output file name bb...bb: Errno (decimal number) cc...cc: Errno string Action Remove the error by referencing ErrorCode (errno).
Message ID Message Text Explanation the HDLM Configuration on page 7-16. KAPL10397 Multiple instances of the dlmcfgmgr Details -E utility cannot be executed The dlmcfgmgr utility for concurrently. Wait for the program managing the HDLM configuration to complete execution and then run cannot be simultaneously the command. executed for multiple processes. Action Wait until the dlmcfgmgr utility ends, and then re-execute the utility.
Message ID Message Text KAPL10703 This version of the OS is not -E supported. Explanation Details The dlmvxexclude utility for assisting creation of the VxVM configuration file was executed on a non-supporting OS. Action It is not necessary to execute this command. For details on the dlmvxexclude utility, see The dlmvxexclude Utility for Assisting Creation of the VxVM Configuration File on page 7-30. KAPL10704 dlmvxexclude completed -I successfully.
Message ID Message Text Explanation It is not necessary to execute this command. For details on the dlmvxexclude utility, see The dlmvxexclude Utility for Assisting Creation of the VxVM Configuration File on page 7-30. KAPL10708 No authority to execute -E dlmvxexclude. Details You do not have permission to execute the dlmvxexclude utility for assisting creation of the VxVM configuration file. Action Execute by a user with root permissions.
Message ID Message Text Explanation Check the existing configuration definition file (/kernel/drv/ aa...aa). KAPL10712 The HDLM support disk information -E does not exist. Details A valid entry does not exist in the /kernel/drv/dlmfdrv.conf file. Action Check the system configuration and then execute the dlmsetconf utility for creating the HDLM driver configuration definition file. KAPL10713 The aa...aa directory could not be -E opened. Errno = bb...bb (cc...
Message ID Message Text Explanation Creation of the VxVM Configuration File on page 7-30. KAPL10716 aa...aa specified by the -d option is -E not a directory. Details The specified file name is not the name of a directory. aa...aa: Directory name Action Specify the correct directory. KAPL10717 The directory specified by the -d -E option does not exist. Details A non-existing directory was specified. Action Specify the correct directory.
Message ID Message Text Explanation Creation of the VxVM Configuration File on page 7-30. KAPL10721 The entered value is invalid. The -E processing will now stop. Details Processing was stopped because an incorrect value was entered three or more times for the message confirming execution of the dlmvxexclude utility for assisting creation of the VxVM configuration file. Action Retry dlmvxexclude command.
Message ID Message Text KAPL10752 The dlmsetboot utility completed -I successfully. End time = aa...aa Explanation Details The dlmsetboot utility completed successfully. aa...aa: year (4 digits)/month/ date hour:minute:second Action None. KAPL10753 The dlmsetboot utility terminated -E unsuccessfully. End time = aa...aa Details The dlmsetboot utility terminated unsuccessfully. aa...
Message ID Message Text Explanation now stop. Execute the utility again. End time = aa...aa or more times for the message confirming execution of the dlmsetboot utility. aa...aa: year (4 digits)/month/ date hour:minute:second Action Re-execute the dlmsetboot utility. For details on the dlmsetboot utility, see The dlmsetboot Utility for Assisting Configuration of an HDLM Boot Disk Environment on page 7-22. KAPL10757 Parameters are invalid.
Message ID Message Text Explanation Action Set the post-migration boot disk device as an HDLM management target. KAPL10761 The file system of the pre-E migration boot disk device is not a UFS file system. Details The file system of the premigration boot disk device is not a UFS file system. Action Re-configure the pre-migration boot disk device file system as a UFS file system. KAPL10762 Do not specify the pre-migration -E boot disk device for the postmigration boot disk device. (aa...
Message ID Message Text Explanation /usr/sbin/ufsdump /usr/sbin/ufsrestore /usr/sbin/eeprom Action After the dlmsetboot utility terminates, check the contents of the messages by the command. If the contents are incorrect, correct them and then try to execute this utility again. KAPL10765 Do you use Sun Cluster or SVM? -I [y/n] Details Do you use Sun Cluster or SVM? [y/n] Action Enter y or n.
Message ID Message Text Explanation bb...bb= backup file name of preupdate configuration file (absolute path) # The following information is output: /etc/system rootdev parameter (only specified) /etc/vfstab /etc/dumpadm.conf DUMPADM_DEVICE=... Action None. KAPL10769 The partition size on the postDetails -W migration boot disk device (aa...aa: The partition size on the postbb...bb) is less than it was before migration boot disk device migration (cc...cc: dd...dd). Are (aa...aa: bb...
Message ID Message Text KAPL10775 The logical device name was -I successfully taken over. (SerialNo = aa...aa, iLu = bb...bb) Explanation Details The logical device name was successfully taken over. aa...aa: SerialNo bb...bb: iLu Action None. KAPL10776 For LU information that is not in the Details -I storage migration information For LUs that are not listed in the definition file (/opt/ storage migration information DynamicLinkManager/config/ definition file, do you want to take dlmmiginfo.
Message ID Message Text Explanation Recreate the storage migration information definition file, and then retry dlmsetconf. KAPL10780 The storage migration information Details -E definition file is invalid. It is empty. The storage migration information (aa...aa) definition file is invalid. It is empty. (aa...aa) file name Action Recreate the storage migration information definition file, and then retry dlmsetconf. KAPL10781 The storage migration information -E definition file is invalid.
Message ID Message Text Explanation Recreate the storage migration information definition file, and then retry dlmsetconf. KAPL10784 The storage migration information -W definition file does not exist. (aa...aa) Details The storage migration information definition file does not exist. aa...aa: file name Action Create the storage migration information definition file, and then retry dlmsetconf. KAPL10785 An error occurred during creation -E of a file used for confirmation. (aa...
Message ID Message Text Explanation If you want to use the LU, reboot the host. KAPL10953 Deletion of the path (aa...aa) -E failed. (Function: bb...bb, ErrorCode = cc...cc (dd...dd)) Details Deletion of a path failed. aa...aa: Filter path name bb...bb: Function name cc...cc: errno dd...dd: errno string Action Contact your HDLM vendor or the maintenance company if there is a maintenance contract for HDLM. KAPL10954 Multiple paths have the same -E property (aa...aa) value for the LU (LDEV = bb...
Message ID Message Text Explanation ¢ ¢ Set LU Load Balance Get Path Status Log Information ¢ Get Local Time ¢ Add Path Information ¢ ¢ Delete Path Information Set Storage Identification Information Action None. KAPL11902 aa...aa has started. PathID = -I bb...bb Details The operation has started on the host. aa...aa: Operation (character string) ¢ Online ¢ Offline bb...bb: The Path ID of the target path (decimal number) Action None. KAPL11903 aa...aa has completed normally.
Message ID Message Text Explanation ¢ Set Storage Identification Information Action None. KAPL11904 aa...aa has completed abnormally. -E Error status = bb...bb Details The operation has completed abnormally on the host. aa...
Message ID Message Text Explanation Execute the DLMgetras utility for collecting HDLM error information, and then contact your HDLM vendor or the maintenance company if you have a maintenance contract for HDLM. For details on the DLMgetras utility, see The DLMgetras Utility for Collecting HDLM Error Information on page 7-3. KAPL11906 GUI information - aa...aa -I Details This information is required for resolving problems. aa...aa: Trace information (character string) Action None.
Message ID Message Text Explanation using the "dlnkmgr set audlog" command or the "dlnkmgr set -audfac" command as necessary. KAPL13602 The audit log configuration file -W cannot be opened. Execute the "dlnkmgr view -sys -audlog" command and check whether a normal result is displayed. Details The audit log configuration file cannot be opened.
Message ID Message Text Explanation Action Contact your HDLM vendor or the maintenance company if there is a maintenance contract for HDLM. KAPL15001 to KAPL16000 Message ID Message Text Explanation KAPL15010 The HDLM utility was executed by -W the user who does not have the authority. Command Line = aa...aa Details KAPL15020 A new LU was successfully added. -I Command Line = aa...aa Details KAPL15021 An attempt to add a new LU has -E failed. Command Line = aa...
Message ID Message Text Explanation KAPL15061 DLMgetras successfully executed. -I Command Line = aa...aa Details KAPL15101 Clear operation was completed -I successfully. Command Line = aa...aa Details aa...aa: Name of the utility that the user executed aa...aa: Command that the user executed KAPL15102 Clear operation has failed. Command Details -E Line = aa...aa aa...aa: Command that the user executed KAPL15103 aa...aa path(s) were successfully -I placed bb...bb. cc...cc path(s) were not.
Message ID Message Text Explanation KAPL15111 The HDLM command was started or Details -W stopped by the user who does not aa...aa: Command that the user have the authority. Command Line = executed aa...aa KAPL15112 I/O information monitoring function -I successfully started. Command Line = aa...aa Details KAPL15113 Could not start the I/O information -E monitoring function. Command Line = aa...aa Details KAPL15114 I/O information monitoring function -I successfully stopped. Command Line = aa...
Message ID Message Text Explanation KAPL15404 The HDLM Manager was executed by -W the user who does not have the authority. - Return Codes for Hitachi Command Suite Common Agent Component When an operation requested of HDLM from Global Link Manager terminates abnormally, or terminates normally with a warning, HDLM outputs one of the return codes described below. Message ID 1002 Explanation Details There is no path on which the operation can be performed.
Message ID Explanation Action Remove the error in the path, and then retry. 1016 Details The target path(s) are already Online. Action Update the host information, and then check the path status. 1017 Details The target path(s) are already Offline(C). Action Update the host information, and then check the path status. 1019 Details An error occurred during HDLM internal processing.
Message ID Explanation Refresh the host information, check the path status, and then perform the operation again. 1027 Details The error monitoring interval and the number of times that the error needs to occur are conflicting with the automatic failback-checking interval. Action Set the intermittent error-monitoring interval to a value that is equal to or greater than (automatic-failback-checkinginterval x number-of-times-error-is-to-occur-forintermittent-error-monitoring).
Message ID Explanation An attempt to recover the path failed because an error occurred on the host. Action Re-execute. If the same error occurs even after removing the reservation, execute the DLMgetras utility for collecting HDLM error information to collect the error information, and then contact your HDLM Refresh the host information, and then perform the operation again.
Message ID Explanation Re-execute the command. If the same error occurs again, execute the DLMgetras utility for collecting HDLM error information to collect the error information, and then contact your HDLM vendor or maintenance company if there is a maintenance contract for HDLM. For details on the DLMgetras utility, see The DLMgetras Utility for Collecting HDLM Error Information on page 7-3. 1036 Details An attempt to acquire information about the HDLM version or SP version has failed.
Message ID Explanation execute the DLMgetras utility for collecting HDLM error information, and then contact your HDLM vendor or maintenance company if you have a maintenance contract for HDLM. For details on the DLMgetras utility, see The DLMgetras Utility for Collecting HDLM Error Information on page 7-3. 1046 Details A parameter is invalid. Action Refresh the host information, and then perform the operation again.
8-134 Messages Hitachi Dynamic Link Manager User Guide for Solaris®
A Sun Cluster 3.2 Commands This section describes the Sun Cluster 3.2 commands. □ Sun Cluster 3.2 Commands Sun Cluster 3.
Sun Cluster 3.2 Commands The table below shows the correspondence between Sun Cluster 3.1 commands and Sun Cluster 3.2 commands used in this manual. For details about the command options, see the documentation for Sun Cluster 3.2.# # Note that there are no differences between Sun Cluster 3.2 commands and Oracle Solaris Cluster 3.3 commands. Table A-1 Correspondence Between Sun Cluster 3.1 Commands and Sun Cluster 3.2 Commands Commands for Sun Cluster 3.1 Commands for Sun Cluster 3.
Commands for Sun Cluster 3.1 Commands for Sun Cluster 3.2 scswitch -F -D device-group-name cldevicegroup offline devicegroup-name scswitch -S -h node-name clnode evacuate node-name #1 Execute this command in step m in Set Up Sun Cluster on page 3-50 of Preparations for a New Installation of HDLM on page 3-45 and in step 14 in Remove Sun Cluster Settings on page 3-180 in Remove Sun Cluster Settings on page 3-180.
A-4 Sun Cluster 3.
B Functional Differences Between Versions of HDLM This section explains the functional differences between newer and older versions of HDLM. □ Functional Differences Between Version 6.1 or Later and Versions Earlier Than 6.1 □ Functional Differences Between Version 6.0 or Later and Versions Earlier Than 6.0 □ Precautions on Differences in Functionality Between HDLM 5.6.1 or Earlier and HDLM 5.6.
Functional Differences Between Version 6.1 or Later and Versions Earlier Than 6.1 In a new installation of HDLM, the default load balancing algorithm has been changed from the Round Robin algorithm to the Extended Least I/Os algorithm. Functional Differences Between Version 6.0 or Later and Versions Earlier Than 6.0 • The format of the displayed HDLM version has been changed when HDLM commands or utilities are executed.
Acronyms and abbreviations The following acronyms and abbreviations might be used in this guide.
D DBMS Database Management System Dev Device DMP Dynamic Multipathing DNS Domain Name Server DR Dynamic Reconfiguration F FC Fibre Channel FC-SP Fibre Channel Security Protocol FO Failover FQDN Fully Qualified Domain Name G GAM Group Membership and Atomic Broadcast GMT Greenwich Mean Time GUI Graphical User Interface # A B C D E F G H I J K L M N O P Q R S Acronyms-2 Hitachi Dynamic Link Manager User Guide for Solaris® T U V W X Y Z
H HBA Host Bus Adapter HDev Host Device HLU Host Logical Unit HTTP Hypertext Transfer Protocol I I/O Input/Output IP Internet Protocol iSCSI Internet Small Computer System Interface L LAN Local Area Network LDAP Lightweight Directory Access Protocol LDEV Logical Device LDoms Logical Domains LLT Low Latency Transport # A B C D E F G H I J K L M N O P Q R S T U V W X Y Z Acronyms-3 Hitachi Dynamic Link Manager User Guide for Solaris®
LU Logical Unit LUN Logical Unit Number M MP Maintenance Pack N NAS Network Attached Storage NFS Network File System NTP Network Time Protocol O OBP OpenBoot PROM OS Operating System P P Port R RADIUS Remote Authentication Dial in User Service # A B C D E F G H I J K L M N O P Q R S Acronyms-4 Hitachi Dynamic Link Manager User Guide for Solaris® T U V W X Y Z
S SAN Storage Area Network SCSI Small Computer System Interface SLPR Storage Logical Partition SMTP Simple Mail Transfer Protocol SNMP Simple Network Management Protocol SP Service Pack SSL Secure Sockets Layer SVP Service Processor U UFS UNIX File System V VTOC Volume Table Of Contents W WWN World Wide Name # A B C D E F G H I J K L M N O P Q R S T U V W X Y Z Acronyms-5 Hitachi Dynamic Link Manager User Guide for Solaris®
Z ZFS Zettabyte File System # A B C D E F G H I J K L M N O P Q R S Acronyms-6 Hitachi Dynamic Link Manager User Guide for Solaris® T U V W X Y Z
Glossary This glossary explains the terms used in this manual. A automatic failback A function for checking the status of paths at regular intervals, and automatically placing the status of a path recovered from an error into the Online status. If a path with an Offline(E) or Online(E) status recovers from an error, an automatic failback will place the path Online.
CLPR(Cache Logical Partition) A function supported by the Hitachi USP series, Universal Storage Platform V/VM series, Virtual Storage Platform series, Hitachi AMS2000/AMS/WMS/SMS series, HUS100 series, and HUS VM for logically splitting up a cache. This function can split up a cache into parity groups in the storage system, so that other parity groups do not affect the cache performance.
F failback A function for placing the status of a path recovered from an error into the Online status, and then switching the access path to the path that was just recovered. failover A function for switching to another normal path if there is an error in the current access path, thereby enabling the system to continue to operate. FCode A program written by the interpreter language that is not dependent on the machine. The HBA that can connect to a boot disk might incorporate FCode for use.
host A generic term for both servers and clients. host device A logical area in a host LU. (See also: host LU, host device name) host device name A name assigned to a host device. A name of the logical device file for an HDLM device is assigned to a host device name. A host device name is assigned using the cUtXdY format.
- The Least Blocks algorithm - The Extended Least Blocks algorithm local boot disk environment An environment in which a boot disk is located on the host. logical device file Device files under /dev/dsk and /dev/rdsk. Note that the term logical device used here is not an LDEV. LU (Logical Unit) A logical unit that is a logical volume defined in the storage system, and with which the host can perform input or output operations. (See also: host LU) N node A server in a cluster.
owner path A path that passes through an owner controller for an LU in the Thunder 9500V Series, or Hitachi AMS/WMS series is in use. A path that passes through a controller other than an owner controller is called a non-owner path. When the Lightning 9900 series, Lightning 9900V series, Hitachi USP series, Universal Storage Platform V/VM series, Virtual Storage Platform series, Hitachi AMS2000 series#, Hitachi SMS series#, HUS100 series#, or HUS VM is in use, all paths are owner paths.
an LU that has been reserved is given only to the host that issued the reservation, so the LU cannot be accessed from multiple paths (coming from multiple hosts) simultaneously. However, because of this, load balancing is not possible. S SAN (Storage Area Network) A high-speed network connecting hosts to storage systems. This network is independent of a LAN and is dedicated to data transfers.
# A B C D E F G H I J K L M N O P Q R S Glossary-8 Hitachi Dynamic Link Manager User Guide for Solaris® T U V W X Y Z
Index Symbols removal 3-187 removing 3-127 upgrading Solaris 4-4 buffer size 3-144 changing the buffer size per monitoring interval duration 3-146 bus number 6-52 /etc/system file setup 3-49 /kernel/drv/sd.
online 6-11 set 6-17 view 6-32 commands of HDLM (dlnkmgr) help 6-4 monitor 6-68 components HDLM system components 2-4 core logic component 2-8 Critical 2-32 cUtXdYsZ 2-10 D default and recommended values 6-18 Dev 2-3, 2-4, Glossary-2 Dev number Glossary-2 directory to which collected information is output 7-4 disk group creating 3-151 disk label 3-45 Displayed program information 6-35 displaying information 6-32 LU information 6-53 path information 6-33, 6-41 path information (by selecting a display item)
G getras.tar.Z 7-6 Global Link Manager 2-41 glossary Glossary-1 H HAM environment Glossary-3 handling path error 5-3 HBA 2-4 HBA (Host Bus Adapter) Glossary-3 hbsa.tar.
preparations for a new installation of HDLM 3-45 preparations for an upgrade installation or reinstallation of HDLM 3-73 installation-information settings file contents 7-32 INSTALLATION_SETTINGS 7-32 installhdlm 7-32 installhdlm.
path information 5-5 program information 5-8 offline 2-16, 6-6 offline(C) status 2-20 offline(E) status 2-20 online 2-16, 6-11 online status 2-19 online(D) status 2-20 online(E) status 2-19 placing online path 5-5 port 2-3 position of the HDLM driver and HDLM device 2-8 product ID 6-48, 6-51 Product ID displayed by the view -path operation 6-52 program error what to do for 5-7, 5-8 online(S) status 2-20 operating environment 6-17 setting 6-17 view [view operation] 6-39 operation name 8-3 operations 6-2 di
Sun Cluster 3-50 VxVM (before installing HDLM) 3-50 single-user mode operation 4-4 SLPR (Storage Logical Partition) Glossary-7 starting HDLM manager 4-15 Solaris in single-user mode 4-3 statistics 6-3 assisting creation of the VxVM configuration file 7-30 assisting HDLM transitions 7-20 creating the HDLM driver configuration definition file 7-23 for collecting HDLM error information 7-3 HDLM component installation utility 7-20 installing HDLM 7-32 installux.
Hitachi Dynamic Link Manager User Guide for Solaris®
Hitachi Data Systems Corporate Headquarters 2845 Lafayette Street Santa Clara, California 95050-2639 U.S.A. www.hds.com Regional Contact Information Americas +1 408 970 1000 info@hds.com Europe, Middle East, and Africa +44 (0)1753 618000 info.emea@hds.com Asia Pacific +852 3189 7900 hds.marketing.apac@hds.
