HP Matrix Operating Environment 7.3 Release Notes Abstract This document describes the HP Matrix Operating Environment 7.3 release content and provides information about limitations, and major and minor issues with suggested actions.
© Copyright 2006–2013 Hewlett-Packard Development Company, L.P. Confidential computer software. Valid license from HP required for possession, use or copying. Consistent with FAR 12.211 and 12.212, Commercial Computer Software, Computer Software Documentation, and Technical Data for Commercial Items are licensed to the U.S. Government under vendor's standard commercial license. Confidential computer software. Valid license from HP required for possession, use or copying. Consistent with FAR 12.211 and 12.
Contents 1 Description................................................................................................5 Update recommendation...........................................................................................................5 Supersedes..............................................................................................................................5 Operating systems and devices...............................................................................................
1 Description HP Matrix Operating Environment is advanced infrastructure management software that allows you to analyze and optimize both physical and virtual resources in the same way. Matrix OE gives you a powerful toolkit to accelerate complex technology projects and simplify daily operations. It enables you to plan capacity and power continuously, provision infrastructure consistently, and balance across physical and virtual resources.
◦ OpenVMS 8.4 ◦ Ubuntu 12.04.3 (HP SIM only) ◦ Microsoft Hyper-V Server 2012 R2 • Hardware support for HP ProLiant ML350e Gen8 v2 • Browser support for: • • ◦ Microsoft Internet Explorer 11 ◦ Mozilla Firefox 24 ESR Storage support for: ◦ HP MSA 2040 iSCSI ◦ HP MSA 2040 SAS ◦ HP 3PAR OS 3.1.2 MU3 SQL Server 2012 AlwaysOn Availability Groups NOTE: Data synchronization to any secondary replicas needs to be disabled during upgrade in order for the upgrade to be successful.
IPv4 for dual stack. For more information, see the IPv6 Support in Insight Management white paper. ◦ Thin provisioning can be specified when defining disks for a virtual server or adding a new disk to an existing service. ◦ Support for hiding unused FlexNICs from the OS is enabled by default (assuming the VC Domain Group supports it, requiring VC 4.10 or later). It can be controlled for individual logical servers in Matrix OE visualization or disabled entirely using the hpio.properties file (hide.
the acceptance check box in the hpio.properties file using the user.license.agreement properties). The administrator can customize the background and font colors using the Organization Administrator Portal (Organization tab). – By default, the telnet link in the Console Access section has been replaced with SSH and VNC (and the existing RDP link). You can configure the desired URL launches for console access (including telnet if desired) using the custom.console.link properties in the hpio.
NOTE: For Matrix OE API enhancement details, see “Matrix OE API change details” (page 61). – Create and configure a Matrix infrastructure orchestration network (createSubnet, deleteSubnet).
Considerations • Matrix infrastructure orchestration ESXi cluster provisioning does not support ESX host creation configured with VXLAN. Matrix OE can create ESXi hosts without VXLAN, and manual post configuration can be done to enable VXLAN in the provisioned server. • Matrix infrastructure orchestration displays all the networks which are present in the vCenter datacenter (including VXLANs from different dvSwitches which are part of different VXLAN scopes or not part of the VXLAN network scope).
1. 2. Select Help→For HP Systems Insight Manager from the top menu. In your browser address bar (if enabled), change the URL from the first line to the second line: https://{CMS-FQDN}:50000/mxhelp/mxportal/en/help-welcome.html https://{CMS-FQDN}:50000/mxhelp/vsemgmt/en/licenses.html 3. where {CMS-FQDN} is the fully qualified domain name of the CMS. Press Enter to display the Matrix OE license management topic. From there, you can select any of the help topics that apply to Matrix OE software.
Support for a rolling upgrade of a CMS federation To decrease federation downtime, Matrix OE infrastructure orchestration 7.3 includes support for a rolling upgrade of a federated CMS environment. A rolling upgrade temporarily allows a federation containing CMSs running different versions of Matrix OE during the upgrade process. IMPORTANT: The primary CMS must be upgraded first. During a rolling upgrade, a CMS running Matrix OE 7.3 supports basic operations against a secondary CMS running 7.2. Not all 7.
9. Open the hpio-loadbalancer-f5 folder and replace the existing files using the files from the desktop of the CMS. 10. Open the hpio-loadbalancer-haproxyfolder and replace the existing files using the files from the desktop of the CMS. 11. Open the hpio-oo folder and replace the existing files using the files from the desktop of the CMS. 12. Restart the OO RSJRAS and RSCentral services. Deactivated logical servers are powered off after upgrade from 6.0 and 6.
3. Orchestration Studio system (C:\Program Files\HP\Operations Orchestration\ Studio\conf\rc_keystore). Restart Operations Orchestration Studio. Certificate key mismatch after Operations Orchestration Studio upgrade After upgrading Operations Orchestration Studio from version 7.51 to 9.x, Operations Orchestration Central and Studio might have different certificate keys. If Operations Orchestration Studio cannot access the URL https://localhost:16443. To resolve this issue, perform the following steps: 1.
Patch required for proper management of Microsoft Windows Server 2008 with Hyper-V systems A patch from Microsoft is required for managing Microsoft Windows Server 2008 with Hyper-V systems. Patches are installed on the Hyper-V hosts, and after you do so, you must install or update Hyper-V Integration Services from the updated host to each VM guest. Do this as a precaution because not all patches affect Integration Services on the guests.
7. 8. IP that is registered with HP Insight Control virtual machine management, or that will be used for registration with virtual machine management. In the pop-up that displays the virtual switch properties, select the service console port and click the Edit button. Change the Network Label to Management Network and click OK. Close the virtual switch properties pop-up window. It can take a maximum of five minutes for the cache to be refreshed and the API to report the ESXi host.
2 Issues and suggested actions The issues and limitations in this release are listed in this chapter, organized by Matrix Operating Environment feature or functional area. If a feature or functional area has no limitations or issues, that category will not appear.
Paper in the Insight Management Information Library at http://www.hp.com/go/ insightmanagement/docs. Configuring Matrix OE with a remote Microsoft SQL Server database Because Windows Single Sign On is used by Matrix OE to connect to the SQL Server database, the credentials associated with the user account executing a command or service are used to connect to the remote database.
Capacity Advisor Limitations Capacity Advisor provides partial support for the HP ProLiant WS460c Gen8 workstation Capacity Advisor does not support the Customize system from hardware catalog feature for this workstation. The system template will not be available in the Edit Scenario - Create System screen.
Infrastructure orchestration Infrastructure orchestration runs on an x64 Windows CMS. Limitations IO/OO Load Balancer workflows not able to process large HAProxy configuration requests The Create Logical Load Balancer Groups workflow targeting the Load Balancer HAProxy may fail with the following error message: The execution of the workflow Create a Logical Load Balancer Groups has failed.
NOTE: VM logical servers manually created from Matrix OE visualization can also specify VM disks on different data stores. • Integrity VM disks must all use the same storage volume names, or be blank. • Cloud server disks do not support storage volume names and must be blank. • Hyper-V VM Linked Clone disks must have the same storage volume names, or be blank. They cannot be spread across different datastores. • Microsoft does not support VM Linked Clone using multiple data stores.
3. 4. Use the HBA vendor's utility to configure the HBA's boot path to boot from the desired SAN boot LUN. Power off the target server after the Boot Controller Order and the HBA's boot path have been configured correctly. Required HBA firmware and OS drivers to access multi-initiator N-Port ID Virtualization (NPIV) data disks The HP blade multi-initiator NPIV feature allows you to configure multiple WWNs for a single HBA connected with a Virtual Connect Fibre Channel module.
Install Microsoft Sysprep tool for customization of Windows 2003 guests on Hyper-V after installation The Microsoft System Preparation (Sysprep) tool, which is required for customization of Windows 2003 guests on Microsoft Hyper-V, is not retained on the CMS after upgrading from previous versions of Insight Software. Without these tools, customization of Windows 2003 guests on Hyper-V results in failures.
Cluster shared volume is not correctly recognized unless all VM hosts in the Hyper-V cluster are registered in Insight Control virtual machine management When only one VM host in a Hyper-V cluster is managed by Insight Control virtual machine management, the following areas are impacted: • Matrix infrastructure orchestration — IO treats a cluster shared volume (CSV) as a cluster disk, and allocates only one VM to the cluster disk.
Suggested action Do not modify the Systems Insight Manager default port of 50000 in the JBoss container. Customizable allocation process deprecated The customizable allocation process functionality in Matrix Infrastructure Orchestration was deprecated in the 7.2 release. The functionality will be removed in a future version.
Doing so will cause the Matrix OE post-install customization step to fail with an error stating Error customizing Logical Server Logical server name realized by target UUID. Suggested action Matrix OE environments can have the newly provisioned server join the domain via an OO workflow. NOTE: When utilizing the PowerShell PSExec command, the OO RSJRAS service should be run as administrator. Minor issues IO and LSM lose DAS disk configurations after an upgrade to Insight Management 7.
Suggested Action 1. 2. Perform manual cleanup, see the Manual clean-up process (physical) in the HP Matrix Operating Environment Infrastructure Orchestration User Guide at http://www.hp.com/go/matrixoe/ docs. Refer to Manual cleanup for HP Insight Control server provisioning. Recreate the service request in Matrix infrastructure orchestration. Create Service request with Insight Control server provisioning fails when deploying RHEL 5.9 OS The installation of RHEL 5.
Before you start a delete service request from infrastructure orchestration: 1. 2. Make sure that the Integrity service was successfully provisioned (the Integrity service is running). Discover the Integrity blade where the Integrity service is running using its IP address from Systems Insight Manager. In a federated CMS environment, discover the Integrity blade on the primary CMS using its IP address from Systems Insight Manager.
1. 2. Edit the install.cmd file, found under the \rdp\ deployment server\hpfeatures\hpagents-ws\components folder. Remove the following two lines: call :installsc wbem%PROCESSOR_ARCHITECTURE% if %cpresult% geq 1 set /a cpresults=cpresults+1 Provisioning a service may fail when storage volumes in a logical server group contain different tags In an IO template, different storage tags can be assigned to the individual volumes in a logical server group.
VM cluster provisioning requires restart of VMware Management Agents for preferred hostname to be registered in DDNS The VM cluster Operations Orchestration workflows enable Dynamic DNS integration by setting a VMware property. For this property to be activated, and the DDNS registration performed, the management network on the ESX host must be manually restarted.
1. 2. On the Firefox “This Connection is Untrusted” screen, click to expand Technical Details or I Understand the Risks. Click Add exception or or you can add an exception. If there is no option to add an exception: a. Open a new tab in Firefox. b. Type the address for the IO self-service portal, https://[your-CMS-name-or-IP]:51443/hpio/portal or https://[your-CMS-name-or-IP]:51443/hpio/portal.jsp. c. Firefox will display the “This connection is Untrusted” screen with the option to add an exception. 3. 4.
Suggested action Before executing archive databasedelete, consult with your database administrator. If you have already executed the command, ask the administrator about performing a shrink operation on the transaction log.
NOTE: HP Integrity Superdome 2 systems obtain monitoring data by executing commands using sudo. Monitoring data may not be available in this scenario. Suggested action Configure HP-UX managed nodes so that no extraneous output is generated when the sudo command is executed.
◦ Not supported: SPM catalog storage volumes. In this release, catalog storage pool entries are supported with Virtual Connect logical servers only. Catalog storage pool entries cannot be used with cross-technology logical servers that may be activated on an ESX or Integrity VM host. ◦ Integrity VM storage requirements: – SAN-based logical server storage is supported, including storage pool entries. File based storage is not supported with Integrity VM cross-technology logical servers.
VM host operation you want to perform Steps to perform on VM guests Steps to perform for the VM host after performing the steps on the VM guests Delete the VM host logical • If you no longer need the VM 1. Unregister the VM host from Insight Control virtual server machine management. From HP SIM, select guests, delete them. Otherwise, Configure→Virtual Machine→Unregister Virtual move all VM guests to a suitable Machine Host… host. 2. Delete the VM host logical server.
• Hyper-V VM linked clone disks must either specify the same storage volume name not specify a name at all. • Microsoft does not support VM Linked Clone using multiple data stores. NOTE: By default, specifying storage volume names when adding a disk to a VM is restricted to administrators. To enable all users to specify storage volume names when adding a disk to a VM, change the value of the hpio.properties user.disk.add.storage.volume.names.restricted properly from true to false.
visualization and imported as a logical server, it is possible that the boot disk controller type is SCSI. In this case, you must install the SCSI controller driver to allow installation of the operating system. If a virtual machine is provisioned by Matrix OE infrastructure orchestration (using either a SCSI or an IDE controller) and the logical server is imported, the virtual machine has an operating system installed so the SCSI driver is not needed.
Shell> drvcfg Configurable Components Drv[2C] Ctrl[29] Lang[eng] 2. Enumerate all SCSI and FC LUNs using drvcfg -s . (drvnum and ctrlnum are displayed as part of step 1.) Press Y for confirmation on enumerating both SCSI and FC LUNs. Shell> drvcfg -s 2C 29 Set Configuration Options Drv[2C] Ctrl[29] Lang[eng] HP AVIO Stor Driver Configuration ================================== Warning: Enumerating all SCSI or FC LUNs increases initialization times.
6. 7. 8. 9. Select the line with the long name and click Remove. Click Add, enter Matrix_Administrators and click Search. If no results are returned enter HPIO_Administrators. Select the resulting group and click OK. Repeat Step 3 but select the Matrix Architect link, then follow Step 4, Step 5 and Step 6. For Step 7 click Add, enter Matrix Designers and click search. Repeat the search and enter HPIO_Architects. Repeat Step 3 but select the Matrix User link, then follow Step 4, Step 5 and Step 6.
4. name that you made note of in Step 1). Issue this command once for each of the WWNs associated with the unmanaged Integrity VM logical server. Create a placeholder logical server to prevent the MAC addresses from being assigned to new logical servers: 1. Create a new logical server template without specifying any storage entries in the storage configuration. 2. In the network selection page, click Add Network Entries. NOTE: LSM shows the MAC addresses that were part of the unmanaged logical server. 3.
wrapper.java.additional.2=-XX:PermSize=32m -XX:MaxPermSize=128m 3. Restart the HP Logical Server Automation service. Minor issues VM logical server may not immediately activate when created using the logical server copy feature When using Logical Server Copy (Tools→Logical Servers→Copy) to make a copy of a virtual logical server with storage, the newly created logical server may not immediately activate.
Suggested action At the Integrity Virtual Machine EFI prompt, execute the following commands to see LUN mappings. When these steps are complete, the user can select the required LUN from the displayed device list to boot up or to install the operating system. 1. Check configurable controllers using the drvcfg command. Shell> drvcfg Configurable Components Drv[2C] Ctrl[29] Lang[eng] 2. Enumerate all SCSI and FC LUNs using drvcfg -s . drvnum and ctrlnum are displayed as part of Step 1.
NOTE: An IO service or a logical server intended to be DR protected using Matrix recovery management must contain its disks and VM files in a single datastore (multiple datastores are not supported). Process Resource Manager Limitations Creating and modifying configuration files with privilege elevation enabled is not supported If privilege elevation is enabled in Systems Insight Manager, you can monitor Process Resource Manager (PRM), but you may experience problems performing PRM operations.
operational details, see the technical white paper on SPM adapter for EMC Symmetrix VMAX at http://www.hp.com/go/matrixoe/docs. • Creating a new volume with a given RAID requires pre-configuration of a storage pool setting instance by the EMC storage administrator who can create a volume of that RAID through EMC native tools (SMC or SYMCLI).
support for port groups: Storage controller port load balancing using SPM white paper at http:// www.hp.com/go/matrixoe/docs. HP 3PAR StoreServe Storage Zero Detect SPM does not recognize the Zero Detect property of volumes on 3PAR StoreServ Storage. SPM 2.3 (included with Matrix 7.3) cannot create volumes with the Zero Detect property. SPM can import and use pre-provisioned volumes with the Zero Detect property, but it does not recognize the property value.
5. 6. 7. Deactivate the logical server. Modify the storage pool entry to make the boot volume redundant. Reactivate the logical server. Matrix KVM Private Cloud Limitations Matrix KVM Private Cloud is available in US English only The current version of Matrix KVM Private Cloud is available in US English only. Suggestion action Specify only US English characters in the Matrix KVM Admin Console text fields.
Major issues RHEL 6.4 with KVM OS Build Plan fails with error The default Insight Control server provisioning RHEL 6.4 KVM OS build plan configures the “br0” device, causing post-installation customization to fail with the following error: Error customizing Logical Server realized by . Suggested action To prevent this, edit the kickstart file referenced in the OS Build Plan as follows. 1. 2.
Intermittent timeouts may occur during VM deployment Deploying a VM to a KVM host using Matrix infrastructure orchestration may occasionally time out with the error: The request for HP Matrix KVM Private Cloud server has failed. … Read timed out connecting to POST. When this occurs, all hosts on the Matrix KVM Admin Console Hosts screen display an error status. Suggested action 1. 2. Retry the create service request in Matrix infrastructure orchestration.
Deleting a large number of virtual machine instances from Matrix infrastructure orchestration occasionally times out Attempting to continuously delete a large number (more than 25) of VM instances from Matrix infrastructure orchestration may not succeed. When this occurs, you may see an error in IO similar to the following: ERROR: Reached TIMEOUT attempting to delete Nova instance with serverId: xxxx-xxxx Failed to stop server. Instance in task_state deleting.
Suggested action Note the number of hosted VM instances when evaluating whether resources on a particular host are available. If sufficient resources are not available, perform one of the following tasks: • Add compute resources to the host. • Free space on the host by deleting existing instances that are no longer needed.
1. Enable the Skip OS Customization check box Set allow.skip.os.customization = true in hpio.properties. This enables a per-server-group Skip OS Customization check box to appear in Designer. 2. Select the Skip OS Customization check box In Designer, navigate to Configure Server Group→Software→1. Select Operating System. Select the Skip OS Customization check box. 3. Ensure appropriate OS customization is done (perhaps via an Operations Orchestration workflow attached to the service template).
Suggested action 1. 2. 3. 4. Create the service from the IO template. Power off the virtual server where the NPIV data disk will be added by selecting Power off servers from the Actions or Server Actions list in IO. Add the required NPIV data disk to the virtual server by selecting Add disks from the Actions or Server Actions list in IO. Power on the virtual server by selecting Power on servers from the Actions or Server Actions list in IO.
To determine if this error is caused by an EFI disk read error, open a console window on the host and look for a message similar to the following: Automatic boot entry execution will start in 1 second . Booting VC Pri. Fibre(0x5001438004C8E7B8,0x1000000000000) Boot Failed. VC Pri. Fibre(0x5001438004C8E7B8,0x1000000000000) Booting HP-UX Primary Boot: 0/0/0/3/0/0/2.0x5001438004c8e7b9.
NOTE: The hpio.properties resides under IOinstalledDir/conf. The lsa.properties resides under VSEinstalledDir/conf/lsa. Suggested action 1. Measure the time required for the power cycle of the Integrity blade being provisioned. NOTE: The time varies based on the blade model, its memory size, and the number of QLogic cards. You can measure the power cycle time from the Integrity blade remote console using Onboard Administrator (OA) or VCEM. 2. Set the serverboot.physical.wait.
3 Support and other resources Contacting HP Information to collect before contacting HP Be sure to have the following information available before you contact HP: • Software product name • Hardware product model number • Operating system type and version • Applicable error message • Third-party hardware or software • Technical support registration number (if applicable) How to contact HP Use the following methods to contact HP technical support: • See the Contact HP Worldwide website for contac
After creating your profile and linking your support agreements to your profile, see the Software Updates and Licensing portal at http://www.hp.com/go/hpsoftwareupdatesupport to obtain software, documentation, and license updates. Warranty information HP will replace defective delivery media for a period of 90 days from the date of purchase. This warranty applies to all Insight Management software products.
For more information about Matrix OE and related products and solutions, visit the following HP websites: • • HP Insight Management: ◦ http://www.hp.com/go/insightmanagement ◦ http://www.hp.com/go/insightmanagement/docs HP Matrix Operating Environment: ◦ http://www.hp.com/go/matrixoe ◦ http://www.hp.
4 Documentation feedback HP is committed to providing documentation that meets your needs. To help us improve the documentation, send any errors, suggestions, or comments to Documentation Feedback (docsfeedback@hp.com). Include the document title and part number, version number, or the URL when submitting your feedback.
A Matrix OE API change details This appendix provides additional Matrix OE API details. See the Web Services Description Language (WSDL) for the formal specification. New and changed Matrix OE API data types The following data types are added: Table 2 CostUpdate Data type Description String serviceEntityName Entity name for cost update. ResourceCostTypeEnum type Resource type for cost update. Float value costupdate value.
New and changed Matrix OE API requests AddNetworkInterfaceToLogicalServerGroup Enhancements to the AddNetworkInterfaceToLogicalServerGroup request include: • The existing tags parameter can now be used to specify storage tags for virtual disks. • The following new optional parameter is added to the end of the parameter list: Name Parameters StorageProvisioningTypeEnum storageProvisioningType Optional. The Provisioning Type for a virtual disk. Defaults to Unspecified.
Boolean redundant Optional. Is true if the when a redundant network interface is to be created. Default value is false. Returns String requestId Request ID for the operation. Exceptions InvalidParameterException N/A SubnetNotFoundException N/A LogicalServerGroupNotFoundException N/A ServiceNotFoundException N/A IOSoapException N/A createSubnet The new createSubnet request enables you to programmatically define subnets. Name createSubnet Defines a subnet from the specified subnetDetails.
editLogicalServers The existing editLogicalServers request is enhanced to support an optional parameter, at the end of the existing parameter list, to control whether or not a reboot is allowed. Optional. Indicates whether the server can be rebooted if necessary to complete the operation. For example, a reboot is required when changing the CPU or memory size of a VMware server when the “hot add” option is not supported or enabled.
runCustomServerAction The new runCustomServerAction request executes the Operations Orchestration workflow associated with a custom server action configured in conf/hpio.properties. Table 8 runCustomServerAction Name runCustomServerAction Runs the Operations Orchestration workflow associated with the specified action. Parameters String serviceName The ID or name of service. String workflowPath (Optional) The path to the Operations Orchestration workflow.
Chargeback API Changes This release adds a new Chargeback API endpoint (V3) to expose isInCloud information for IP addresses in chargeback reports. This information was not exposed in the previous endpoints.
B HP Capacity Advisor enhancement details Systems with a possible infinite loop locates systems with possible resource problems due to a profile indicating that an application is in an infinite loop. These systems have had at least one CPU core fully utilized for more than 24 hours, or at least half of one hyperthreaded CPU core occupied for more than 48 hours. The following new attributes are included in the attributes available for use in the analysis queries on the CPU tab.
