Prepare to upgrade the partitioned server

Prepare for the upgrade

Obtain and review the preventive service planning (PSP) information.
The PSP document for V5R4 upgrades is SF98168. Review this document for the most current information about conditions that can affect your upgrade.
Review the Memo to Users.
This document provides information about software and hardware changes that can potentially affect your system operations.
View the Migration Web site.
This Web site includes information about the current migration and upgrade paths, a link to relevant Redbooks™, and information about converting from SPD to PCI-attached I/O.
Review the Performance Capabilities Reference available on the Performance Management Web site.
This reference provides information about server performance that is useful for customers who plan to upgrade their servers and software.
When your order arrives, contact the service representative and schedule the server upgrade.

Prepare the source server

Ensure that you have installed the latest resave and cumulative fix (PTF) package of i5/OS™ V5R4 on all partitions of the source server.

For information about the resave and cumulative fix (PTF) package, see Info APAR II13365 on the iSeries™ and AS/400^® Technical Support Web site. For information about installing i5/OS, see Upgrade to i5/OS Version 5 Release 4.

Install fixes (program temporary fixes or PTFs) on all partitions of the source server. Use the information in this topic to put fixes on all partitions of the source server before upgrading to the target server.
Verify that all applications are running.
Perform device configuration cleanup as follows:
- Use the Work with Line Descriptions command (WRKLIND) to view and delete unused lines and line descriptions for all partitions.
- Use the Work with Device Descriptions command (WRKDEVD) to delete unused device descriptions for all partitions.
Note: You can use these commands to view and verify that all of the descriptions listed are in use. If you are sure one is not in use, delete it. If you are not sure, you should leave it.
Assign all unassigned I/O resources to active partitions.
For information about assigning resources to a partition, see Perform dynamic movement of resources.
If the target server has fewer processors than the source server, you can adjust the configuration on the source server to reflect the target server's processing resources. For information, see Dynamic movement of processing power.
If necessary, you can adjust the minimum memory values per partition. For information, see Dynamic movement of memory.
For each partition, remove failed or nonreporting resources that are no longer associated with existing hardware.
Logically remove any configured disk units that will be removed during the upgrade, such as disk units in SPD-based expansion units that will be removed (not converted) during the server upgrade. For information, see "How to Remove a Disk Unit from an Auxiliary Storage Pool" in Chapter 19. "Working with Auxiliary Storage Pools" of the Backup and Recovery Guide.
Note: Disk units that are not physically removed may be inadvertently added back into the configuration during an IPL.
Ensure that your console is set up and ready.
Use the Retrieve Configuration Source (RTVCFGSRC) command for each partition to build a source file member for the specified existing line, controller, and device descriptions. You will need this configuration source if any configuration objects must be manually changed or re-created.
Gather performance data.
You can use the performance data you collect from the source server as a base with which to compare the target server's performance after the upgrade is complete.
Save each partition of the server using GO SAVE option 21 on a tape drive that is compatible with the tape drive on the target server.
This is vital for recovery should you experience errors when upgrading the server.

To determine if your tape drives are compatible, see the Storage Solutions Web site.

Document all partitions of the source server

To ensure that the source server is fully documented before the upgrade, perform these tasks:

Use one of the following to document each partition of the source server:
- If your country or region supports the Work with Order Information Files (WRKORDINF) command and if the installed server is capable of using Electronic Customer Support, do the following:
  1. On the command line, enter WRKORDINF.
  2. On the next display, type option 1 (Send to IBM^®) next to QMA nnnnn where nnnnn is the serial number of the server and press Enter.
- If you cannot use the Work with Order Information Files (WRKORDINF) command, do the following:
  1. On the command line, enter DSPSFWRSC *PRINT.
  2. Print the spooled file.
Print the disk configuration status from each partition of the server. Be sure to record the serial number of the load-source disk unit for each partition.
Print the parity set configuration and status of the server from each partition of the server. Be sure to mark each printout for the partition it represents.
You can use this information to analyze and plan for disk unit configuration and protection. It can also be used to correct any problems that occur during the upgrade.
Display, verify, and print the hardware resource information for each partition.
Use the information in this topic to record the hardware resource information so that you can perform the necessary resource management after the upgrade is complete.
From the primary partition, Print the system configuration for logical partitions.
This gives you specific partition information such as system resources, processors, main memory, and any unique system values associated with partitions.
Using the system configuration printout, record the partition owner and serial number of each IOP on Bus 2.
These IOPs may not be on Bus 2 after the upgrade is complete. If not, you can use this information to reassign the IOPs to the correct partitions.
Print the status of all fixes (program temporary fixes or PTFs) for each partition of the source server by entering the following command:
```
DSPPTF LICPGM(*ALL) OUTPUT(*PRINT)
```
You can use this information to verify that you have the most current fixes on the source server.
Record the IPL type and mode of the source server so you can return the server to this mode after the upgrade is complete.
- IPL type = _________
- IPL mode = _________

Prepare for the service representative

Record the user IDs and passwords for the service representative to use.
Gather the following printouts for the service representative:
- System configuration list printouts
- Disk configuration status printouts
- Parity set configuration and status printout
- Device descriptions printout
- Controller descriptions printout
- PTF level printout
- Description label locations printout
- Partition information printouts
- LPAR Validation Tool output
For each secondary partition, record the IPL System Action and then set the IPL System Action to HOLD as follows:
1. From the primary partition, enter STRSST to start system service tools (SST) and sign on to SST.
  Note: To use system service tools, you need a valid service tools user ID.
2. From the SST main menu, select option 5 (Work with system partitions), and press Enter. The Work with System Partitions display appears.
3. Use the Work with System Partitions display to select each secondary partition. For each secondary partition, record the IPL System Action and then set the IPL System Action to HOLD.
On each partition of the server, record the setting for each of the following system values:
- Local controllers and devices (QAUTOCFG) = ___
- Type of restart (QIPLTYPE) = ___
- Automatically adjust memory pools and activity levels (QPFRADJ) = ___
Then change the values of these settings on each partition of the server to the following:
- Local controllers and devices (QAUTOCFG) = Do not automatically configure local controllers and devices (0)
- Type of restart (QIPLTYPE) = Attended, console in debug mode (2)
- Automatically adjust memory pools and activity levels (QPFRADJ) = Periodically after restart (2)
For information about working with system values, see i5/OS system values.
Turn the server over to the service representative to perform the server upgrade. This includes giving the service representative the printouts, the user IDs, and passwords that you have prepared

Now that you have finished preparing for the server upgrade, the next step is to complete the server upgrade.