PTF MH01060

Description

Updates your HMC from HMC V7R3.1.0 to HMC V7R3.2.0.

Package information

Package name

Size (bytes)

Checksum

APAR #

PTF#

HMC_Update_V7R320_1.zip
HMC_Update_V7R320_2.zip
HMC_Update_V7R320_1.iso
HMC_Update_V7R320_2.iso

448843903
587757793
450058240
593747968

21397
28531
55065
30828

MB02134

MH01060

Splash panel information (lshmc -V command output)

Version: 7 Release: 3.2.0
Service Pack: 0
HMC Build level 20071018.1
","base_version=V7R3.2.0

Fixes

The following fixes are addressed in this PTF:

This package updates your HMC from HMC V7R3.1.0 to HMC V7R3.2.0. You can also reference this package by APAR MB02134. All enhancements and fixes included in this package is the same as for PTF MH01059.

Server and Partition Management

  • I/O Reporting Partition (Hardware Discovery). Available only for POWER6 hardware.
    • Provides summary of I/O attached to the system pre-OS for configuration.
    • Provides HMC users more detailed view of slot contents including child devices.
    • Allows administrators to view Ethernet MAC addresses and FibreChannel WWN�s from the HMC before partitions are configured.
    • Provides detailed inventory to System Plan Tool to verify system configuration and validate customer orders
  • Ability to create multiple virtual processor pools, providing easier license management by capping the number of processor cycles that can be consumed by a group of uncapped logical partitions.
  • Partition mobility - a new HMC function, on systems that are capable to use this function, that enables moving a partition from one managed system to another. The active partition mobility function provides the HMC administrator the ability to non-disruptively move a running partition, with its operating system and applications, profiles, and virtualized resources, across two capable and compatible servers that are under the control of the same HMC. The inactive partition mobility function provides the ability to move an inactive, powered-off partition, with its profiles and virtualized resources, across two servers that are under the control of the same HMC.
  • Support for redundant POWER6 FSP failover.

Platform Management

  • Added support to allow the use of Kerberos to authenticate users on HMC. Users are created on the local HMC, but uses a remote Kerberos server for authentication.
  • Added support to setup Kerberos Key Distribution Center (KDC) server configuration.
  • Corrected Manage Dumps task handling of dump parameters for a POWER5 server.
  • Redundant FSP enhancements and corrections.
  • Added a Deactivate Attention LED task that allows a user to deactivate the attention LED's on multiple servers and/or partitions. This new task simplifies the previous methods for deactivating LED's, View System Attention tasks for systems and Manage Attention LED for partitions, which were single target tasks.
  • Corrected a problem where a FSP dump could not be initiated from HMC when system is in poweron state giving unrecoverable error message.
  • Added console log events for adding and deleting NTP servers to create audit trail.
  • Corrected a problem in the HMC backup critical console data task where it does not detect specific permission problem on a remote directory on the ftp server. This problem prevented backup from completing although a message was displayed that indicated the task completed successfully.
  • Corrected errors that caused mail to be sent to the HMC console resulting in performance degradation on the HMC.
  • Corrected an issue with backup of the HMC where their HMC archive data may be directed to an unknown location. If the user did NOT specify an optional directory to offload their HMC archive to, there will be an internal 'cd /' issued from within the ftp session. Where that directs their data depends on how their ftp server is set up. It may be re-directed to user's home directory or to their ftp server's physical "/" directory.
  • Corrected an issue with the HMC interface with multiple column tables where a sort option was not performing correctly.
  • Corrected user interface code to handle the change in formatting of the V7 HMC model type and display the information correctly.
  • Corrected a problem in Systems Management where a user could not toggle between Capacity on Demand and default views using MS Internet Explorer V6.0.
  • Fixed a problem where data replication of 'Group Data' failed to replicate changes to any of the slave HMCs.
  • Corrected a problem where the HMC User Interface Task Bar was flashing continuously.
  • Updated the Network Settings task's reboot/restart requirements. A reboot is no longer required for most modifications to network settings.
  • Enhanced Partition Availability Priority user interface.
  • Miscellaneous improvements to the Guided Setup Install Wizard.
  • Corrected the deployment of the main HMC user interface window to adjust the screen height on the local console.
  • Corrected a problem when using a context menu or drop-down menu in the HMC Work Pane, the opened menu will freeze.
  • Corrected an error found when using the provided icon or drop down options to filter a table to reduce the number of visible rows. Instead of only rows that are NOT filtered out being selected, all rows of the table are selected.
  • Corrected problems seen on an HMC console with more than one console active (local and remote), where tasks were not visible and panels were not fully rendered.
  • Corrected a problem with Network Settings when attempting to return the network interface from an open, non-DHCP Server configuration (on the private subnet shared with the FSPs and DHCP Server) to a previous private, DHCP Server configuration.
  • Corrected a problem in Manage Task and Resource Roles with copying the customized resource roles.
  • Clarified and handled the field 'No IPv4 address' on the Change Network Settings/Adapter Details panel.
  • Enhanced HMC User Roles so that users assigned Super User roles (hmcsuperadmin) will be able to customize their individual user settings independently of other HMC Super Users. Previously all user assigned Super User roles shared user settings.
  • Added help to Customizable Data Replication for "Customer Information" data.
  • Enhanced Change User Password task user interface.

Power Management

  • With power saver mode capable system, customers can enable this feature through HMC to save power by dropping the voltage and frequency to a fixed percentage.
  • Customers can now schedule the power saver mode feature through the schedule operations task by selecting a managed system.
  • IBM Systems Director Active Energy Manager can collect trending information and perform other advanced EnergyScale features by utilizing HMC's network connections to managed systems.
  • For more information on EnergyScale feature, refer to the white paper for IBM EnergyScale for POWER6 Processor-Based Systems.

Service Management

  • Enhanced the HMC Serviceable Event feature to not only provide Serviceable Events for the managed system, but to now also include Serviceable Events for the HMC.
  • SNMP traps were extended to generate HMC related notices.
  • New HMC monitored resources including:
  1. CCFW JVM activity.
  2. Total HMC system usage.
  3. Individual process usage.
  4. Disk partition utilization.
  • A new command, chhmcfs, was created to manage log files.
  • Enhanced ServiceRM to handle increased number of partitions in a single PHYP based system.
  • The design ServiceRM was enhanced to control the flow of data between itself, the RMC daemon and the client on the HMC to better handle memory and alleviate contention for resources.
  • ServiceRM was also modified from handling RMC requests one at a time to handling multiple requests. Previously, if one request did not complete the subsequent requests would be queued and resulting in possible deadlock. Now requests are honored with multiple threads.

Command Line

  • The following commands have been added to manage HMC file system disk space usage:
    • chhmcfs - frees up space in HMC file systems
    • lshmcfs - lists HMC file system disk space usage information
  • A new command, rnvi, has been added to enable HMC users to edit text files in a restricted mode.
  • The following commands have been added to configure and list power management settings for a managed system (POWER6 servers only):
    • chpwrmgmt - changes power management settings
    • lspwrmgmt - lists power management settings
  • The following commands have been added to support partition mobility (POWER6 servers only):
    • lslparmigr - lists partition migration information
    • migrlpar - performs a partition migration operation
  • The following commands have been enhanced to support partition mobility (POWER6 servers only): chsyscfg, lssyscfg, and mksyscfg.
  • The following commands have been added to support Kerberos configuration on the HMC:
    • getfile - gets and deploys the Kerberos service key (keytab) file on the HMC
    • rmfile - removes the Kerberos service key (keytab) file from the HMC
  • The following commands have been enhanced to support Kerberos configuration on the HMC and remote Kerberos authentication: chhmc, lshmc, chhmcusr, lshmcusr, and mkhmcusr.
  • The following commands have been enhanced to support multiple shared processor pools (POWER6 servers only): chsyscfg, lssyscfg, mksyscfg, chhwres, lshwres, and lslparutil.
  • The following commands have been enhanced to support hardware discovery (POWER6 servers only): chsysstate and lshwres.
  • A new option has been added to the chhmc command to set the SSH protocol version for the HMC to use.
  • The chhmc and lshmc commands have been enhanced to support enabling and disabling remote web browser access to the HMC.
  • A new �s� option has been added to the getupgfiles command to allow secure FTP to be used to transfer the upgrade files.
  • The lssyscfg -r sys command has been changed. The sp_failover_enabled and sp_failover_state attributes are no longer output unless the -F option is specified on the command. Now those attribute values will only be displayed if either the -F option is specified with no attribute names, or if the sp_failover_enabled and sp_failover_state attribute names are specified with the -F option.
  • The -o noprobe option for the mksysplan command has been deprecated. It has been replaced by the --noprobe option. In addition, two new options for inventory collection have been added to the mksysplan command: --nohwdisc and --noinvscout.
  • A new option has been added to the sendfile command to allow the user to specify the name of the file on the remote system.
  • A new option has been added to the updlic command to synchronize Licensed Internal Code on redundant service processors, and also on redundant Bulk Power Controllers.
  • Hardware Discovery command line enhancements:
    • chsysstate -r sys will include the new option onhwdisc
    • lssyscfg -r sys will display the new cec capability for Hardware Discovery, hardware_discovery_capable
    • lssyscfg -r lpar will display the new lpar definition state for Hardware Discovery in the lpar_env and lpar_type attribute.
    • lssyscfg -r lpar will display the new VSP definition state for Hardware Discovery in the state attribute.
    • lssyscfg -r prof will display the new lpar definition state for Hardware Discovery in the lpar_env attribute.
    • lshwres -r io -rsubtype will include a new type slotchildren to list Hardware Discovery slot information

 

Licensed Internal Code (LIC) update

  • Enhanced Code Update logic to disallow accept/reject if all components are running on the "wrong" flash side:
    • If all components are on the T side, the reject option is grayed out in the GUI, and the "updlic -o j" command will give the following error message:
      • The reject operation cannot be performed because all components are running on the temporary flash side.
    • Likewise, if all components are on the P side, the accept option is grayed out in the GUI, and the "updlic -o c" command will give the following error message:
      • The accept operation cannot be performed because all components are running on the permanent flash side.
    • If there is a mixture of components on the T and P sides, the accept/reject operations can be performed to the subset of components that are on the correct side for the operation (i.e. those running on T side for accept and those running on P side for reject). In this situation, a message will be displayed to the user asking if they wish to continue.
  • Enhanced estimated time values.
  • Amended Code Update to make E302F8A5 an informational log instead of an error log when exceptions are received from lslic command in an effort to determine HMC code update ownership.
  • Corrected a code update issue where if multiple updates were started at approximately the same time but to different repositories the data from one repository survey could be overwritten by the second.
  • Corrected code update to support updating of systems that have only one BPC.

Scheduled Operations

  • Added support to Scheduled Operations for Utility Capacity on Demand (CoD). There are two different scheduled operations for Utility CoD:
    • Moving Utility CoD processors in/out of the shared processor pool. This will be accomplished via the chcod command.
    • Setting the maximum Utility CoD minute usage limit. The idea is that the customer could schedule the setting of this limit on a monthly basis, therefore implementing a monthly Utility CoD minute usage limit. This will be accomplished via the chcod command.
  • Corrected Scheduled Operations Backup Profile Data to include a --force parameter to always overwrite any pre-existing backup file.
  • Corrected a problem seen in Dynamic configuration Scheduled Operation View Details. This was an issue that only affects Dynamic Reconfiguration Scheduled Operation that were migrated from the 6.1.2 HMC.

Problem Analysis

  • Corrected handling of contention for dumps from 2 HMC's to prevent the incorrect HMC from retrieving the dump and multiple call homes for the same error. A serviceable event will be created on one of the HMC's indicating a dump could not be retrieved.
  • Enhanced Dump Manager to track call home status and problem association.
  • Enhanced SFP Serviceable Event Overview sorting.
  • Modify serviceable event text for PEL-based events to not include PEL severity information.
  • Enhanced call home data to include a reference code summary file to provide a more concise, complete, and readable version of the serviceable event data that includes all serviceable events on the HMC.
  • Amended refcode lookup code to handle the refcode format passed in via the partition view's refcode column that includes the partition as part of the input. (e.g. B2004158 LP=002)
  • Added a busy/processing window to eServer Registration process interface so that it is apparent that background processing is going on.

Service Agent

  • Change to handle 9 character refcodes.
  • Corrected a problem with remote session when running with Firefox browser handling password updates and resets.

Repair and Verify

  • Added support to concurrently add a new GX adapter as an upgrade and for the concurrent (cold) repair support for GX Adapters. The adapter must be deconfigured/garded during IPL/reIPL prior to the repair.
  • Added support for I/O Drawers MTM 7041-SD1, Feature Code 5886, and MTM 7214-1U2. Since these drawers are connected via SAS cables from the I/O adaptors in the server, the HMC will not be able to detect their presence in a configuration. The Repair and Verify procedures will be written to instruct the user to manually interact with the drawer. The procedures for each FRU will be displayed within a browser in HTML format.
  • Corrected the 5094 exchange procedure graphics to include the locations of the connectors on the card in location C10.
  • Corrected repair procedure for PCI adapter cards on System I.
  • Enhanced and/or corrected instructions for Service Processor cable removal
  • Enhanced concurrent repair instructions for removable media devices to include quiescent of applications running on operating systems that may be using a removable media device and using operation procedures to manually shut down of all applications and logical partitions
  • Enhanced process for exchange of VPD passthru card
  • Corrected panel flow for the exchange of the power cables for the 5094
  • Corrected/enhanced concurrent and nonconcurrent procedures for the exchange of MMA fans.
  • Add support to procedures and service guides to handle the I/O expansion units G30, 5790, 5796 and 5296.
  • Clarified the A2 Fan Exchange procedure to instruct the user to replace the Op Panel only if it was removed.
  • Corrected the way that surveillance errors are logged by the HMC for valid managed systems without an MTMS. Previously, by default RV launched Info Center content. Now the correct isolate procedures will be run.
  • Corrected the procedure for an MMA fan to be concurrent.

Known issues in HMC V7R3.2.0

  • Concurrent FRU replacement of Feature Code 1802 GX Dual Port 12X Channel Attached Adapter fails with a Null exception when the operation is performed from an HMC at levels V7R3.2.0 base or V7R3.2.0 with Service Pack 1 installed. The failure will occur during the reconfiguration steps which follow the physical adapter replacement. Recovery of the adapter and it's attached resources will require a platform IPL. This issue is resolved with the installation of PTF MH01100.
  • A timing issue exists where 2 panels, using the same set of information, and one panel modifies some data and the other panel knows nothing about the modifications. This situation can result in "stale" data overwriting newer data. The problem will be fixed in a later PTF.
  • If a HMC is attached to an initialized system without micropartitioning turned on, LPAR profile creation will fail. The workaround is to rebuild the managed system. The problem will be fixed in a later PTF.
  • A DLPAR "Move Memory" operation will partially fail if the move hits pinned memory. Sometimes the pinned memory is higher than the reported minimum. The amount of memory moved will be some fraction of the amount of memory requested in the move. The error message gives the impression that the whole operation failed. The problem will be fixed in a later PTF.
  • If a Virtual Ethernet Adapter is added with a VLAN ID of 1, the adapter will show up as missing in the LPAR Properties panel for an AIX partition. The problem will be fixed in a later PTF.
  • When a customized user logs in, the "tip of the day" panel displays a task error. The problem will be fixed in a later PTF.
  • After performing a dynamic logical partition action on a virtual serial adapter, the virtual adapter properties window may only show the lpar id and value "65536" for the name field. The Manage Profile and dynamic logical partition windows show the name properly. A reboot of the HMC will resolve the inconsistency and populate the name field. The problem will be fixed in a later PTF.
  • After upgrade to V7R3.2.0 if a backup critical console data at that version is later restored during a recovery of the HMC, then the first initial login attempt is not accepted and a second login is required. The problem will be fixed in a later PTF.