SFA OS Release 1.4.1 SFA™ OS Version 2.0.
Important Information Information in this document is subject to change without notice and does not represent a commitment on the part of DataDirect Networks, Inc. No part of this manual may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying and recording, for any purpose other than the purchaser’s personal use without the written permission of DataDirect Networks, Inc. © 2013 DataDirect Networks, Inc. All rights reserved.
TABLE OF CONTENTS 1.0 OVERVIEW .................................................................................................................................................... 4 2.0 INSTALLATION ............................................................................................................................................. 4 2.1 Overview ...........................................................................................................................................................
1.0 Overview This document applies to the SFA OS version 2.0.0.3-17400 release for DataDirect Networks’ SFA12K-40, SFA12K-20, SFA12K-20E, SFA10K-X, and SFA7700 storage systems. It details upgrade procedures, enhancements, resolved issues, known issues with workarounds, and recommended practices associated with SFA OS 2.0.0. “SFA10K” represents SFA10K-X and SFA10K-M. “SFA12K” represents SFA12K-40, SFA12K-20, and SFA12K-20E. 2.0 2.1 Installation Overview SFA OS version 2.0.
2.4 Pre-Firmware-Upgrade Procedure: SFA12K-40 and SFA12K-20 SFA12K-40, SFA12K-20, and SFA12K-20E NOTE: Before upgrading SFA OS on an SFA12K system, please read this section carefully. SFA OS 2.0.0 has a feature that will upgrade the SFA12K Baseboard Management Controller (BMC) firmware version automatically as part of the SFA OS upgrade. The automatic update will only function with SFA12K BMC versions newer than 21.54. Prior to upgrading to 2.0.
Issue 'SHOW ENCLOSURE ALL' to identify your current BMC version. For example, see Figure 2.
4. Issue the command, SHOW UPS * ALL_ATTRIBUTES. You should see the message shown in Figure 4. Battery Mfg. Date: Thu Sep 8 4:10:30 2011 Battery Life Remaining: 730 days Figure 4 3.0 Enhancements SFA OS Version 2.0.0 adds the following enhancements. • Quality Improvements o Ability to mirror system disks on SFA10Ks NOTE: • • • • For more details on adding disks to the mirror set, see section 2.19 of the SFA OS CLUI Command Reference Guide.
using CLUI. The 'app show stack' command now includes this information within the IOC’s output details. o Customer Controller Replacement Support In the event of a controller failure, the process of replacing the failed controller can be now performed by a customer. Details on how to replace a failed controller can be found in the SFA OS User Guide.
Previously, if you issued a subsystem shutdown on the SFA controller, the VMs were shut down without warning. Now, a check is made and if VMs are running, the shutdown command will fail with an error message. This will prevent virtual machines from being shut down by accident.
A controller crash occurred when the shutdown subsystem abrupt command was issued on controller 0 and then again on controller 1. This was due to an issue where the second controller forwarded the second command to the master controller while it was trying to shutdown. • SFA12K controller crash with JEX_Storage_EnqueueRequest This issue was due to a problem in the logging agent inside the SFA OS.
5.0 5.1 Known Issues Common • The drive serial number reported in SHOW PD output may contain additional characters that are not on the disk label. The serial number on the disk label will always be contained within the potentially more verbose string reported by SFA OS. • The SS7000 enclosure does not support the SET SLOT xx POWER OFF command with enclosure firmware versions prior to 05.02.01.
o If the ULA field has the value "0000000000000000", continue to wait and not issue any CLI commands besides "show controller".
• If Installing Older DDN Solutions on SFA12K-20E, Potential Network Ordering Issues May Arise If you install a version of GRIDScaler older than 2.0.0 or EXAScaler older than 1.6.1 and you are using enhanced network management in the VMs (sr-iov virtual function interfaces), you could run in to an Ethernet ordering issue during installation where eth0 may not be associated with the expected driver. This could result in failure of user mode networking (the port 555x access) or other problems.
app app app app app app bind bind bind bind bind bind stack stack stack stack stack stack 32769 32769 32770 32770 32771 32771 ioc ioc ioc ioc ioc ioc 34560 34561 34560 34561 34560 34561 Now, shut down and start up the VMs/stacks again using CLUI. After they boot up, you can configure networking on the virtual machines. • Suggested Ethernet Ordering in Virtual Machines for SFA12K-20E For new installations, DDN recommends the following virtual machine Ethernet ordering starting in SFA OS 2.0.0.
• • • To improve failover time and to prevent I/O errors, the following settings in multipath.conf are recommended: (These settings can be changed for the defaults section which will apply to all devices, or just under the SFA devices): checker_timeout 5 dev_loss_tmo 10 fast_io_fail_tmo 5 These settings are included in the DDN multipath package version 1.5-5 and above.
• In an InfiniBand switch environment, an online upgrade causes an I/O error while upgrading the second controller with SLES 10 SP4. In an FC environment, there are no issues because the discovery rate is much quicker. The only option for SFA upgrades with SLES 10 SP4 is to perform the upgrade offline. To perform the upgrade offline, 1. Issue the CLI command set subsystem offline. 2. Perform the controller firmware upgrade on both controllers. 3. Issue the CLI command shutdown subsystem restart. 4.
Known issues with the SFA7700 are presented in the following tables. Also included are the issues’ frequencies, impact on customers, and resolutions. BIOS: Build 23_001 Issue Freq. Customer Impact Resolution Warm boot hang at step B2 Rare Controller restart required. Manually power-cycle controller by unseating the canister from the enclosure for 60 seconds, then reseat. If this does not turn off the canister, try pressing the hidden power button. Issue Freq.
Issue Restarting a controller does not disable the battery pack. FPGA-27 (4) Freq. Rare Spurious FPGA failure FPGA45 (3) Rare FPGA warm boot hang. FPGA-74 (2) Rare SHUTDOWN does not shut down controllers (continue to run on battery). FPGA-77 (4) ~50% IPMI Power Off does not always turn the canister completely off. FPGA-78 (4) ~30% A long pushbutton push does not force the canister to power off.
5.7 Issue controller instead of holding it in reset FPGA-80 (4) Freq. SEP reset signals false power failure. FPGA-8 100% Customer Impact update the FPGA version reported will be incorrect (0.2550.255) SHUTDOWN ENCLOSURE 1 RESTART results in a false power failure which causes both canisters to save cache and restart. This is only a problem for the head disk enclosure Resolution disk enclosure for 60 seconds. SHUTDOWN SUBSYSTEM and manually power cycle the head disk enclosure. Red Hat 6.
4. opensm -g (where port GUID is the port GUID from ibstat output 5. run_srp_daemon -R 20 -T 10 -nce -i -p (where ca_name is the channel adapter name and port is the port number) 6.0 Recommended Practices The recommended best practices for SFA OS version 2.0.0 include: • SS6000 DEM and I/O module replacement. For SS6000 DEM and I/O module replacement, see the SFA OS User Guide for the procedures. • Reduce host I/O load during background initializations.
8.0 Support To submit debug information to DDN, use the diag tgz command. Diag is a captive account on SFA OS systems that is used for performing operations on a controller when there’s no CLI access or the CLI isn’t appropriate. Perform diag operations by using ssh to login to the diag account. The command syntax is: ssh diag@xxx.xxx.xxx.xxx diag-command (where xxx.xxx.xxx.xxx is the controller’s IP address and diag-commands are listed in Table 1.
World Headquarters 9351 Deering Avenue • Chatsworth, CA 91311 ddn.com Phone: +1.818.700.7600 Fax: +1.818.700.
