This document describes how to confirm if a server can be managed by HMC Version 7.7.8 and later.
Resolving The Problem
HMC uses a storage area on each managed server to persist the server configuration,
primarily PowerVM partition profiles. HMC Release 7.8.0 and later increases the usage of the storage
area by adding another (mostly hidden) profile for each partition. Servers already containing
a large number of profiles may not have sufficient space to allow HMC V7R7.8.0 to run properly.
HMC V7R7.8.0 has checks for sufficient space in this storage area and will stop the connection
process with a Version Mismatch if sufficient space does not exist. Of course, it is preferable to know
if sufficient space exists prior to the upgrade. To allow for automated checking prior to the upgrade, a space evaluation
tool has been provided for release V7R7.7.0 and V7R7.6.0.
Instructions for Version 7 Release 7.7.0 and Release 7.6.0
V7R7.7.0 Apply Service Pack 3 and fix MH01400 or later.
V7R7.6.0 Apply Service Pack 3 and fix MH01384 or later.
To determine if a server can be managed by a HMC V7R7.8.0 and later use the command lsprofspace.
To check a specific server use the command:
lsprofspace -m <managed-system> -F can_upgrade_to_780
To check all servers managed by the HMC use the command:
IFS=',';lssyscfg -r sys -F name | while read mname ; do lsprofspace -m "$mname" -F can_upgrade_to_780 | while read canupgrade ; do echo "$mname" can upgrade: "$canupgrade" ; done; done;
A value of 1 implies the server can be safely upgraded. If one or more servers has a value of 0 then see recovery section below. For further information on using the command type "lsprofspace --help" or "man lsprofspace".
Instructions for Version 7 Release 7.5.0 and earlier
To make a quick estimation, use the following procedure for each server managed by the HMC prior to upgrading.
- count the profiles for all VIOS partitions (for that managed server) and multiply that count by 10.
- Some may find the following piped command useful to obtain this count:
## Count all vios profiles
do lssyscfg -m "$mname" -r prof -F lpar_env | grep vioserver | grep -c . | while read count ; do echo "$mname" vios profiles: "$count" ; done; done;
Don't forget to multiply this count by 10.
- count your client partitions. Add the number of client profiles (across all client partitions).
- Again, the following piped command may be used to obtain this count:
## all nonVIOS profiles
do lssyscfg -m "$mname" -r prof -F lpar_env | grep -v vioserver | grep -c . | while read count ; do echo "$mname" non-vios profiles: "$count" ; done; done;
- For each managed server, add the two numbers together and compare to the limit below corresponding to the managed server machine type and model.
Note that this limit includes a buffer of about 20%.
|Power Server family||Power Server |
Profile calculation limit
(on managed server)
Supporting Code Stream
|Power 5||9119-590, 9119-595, 9406-595, 9118-575|
Size cannot be updated
|9116-561, 9117-570, 9406-570|
|9113-550, 9124-720, 9133-55A, 9406-550|
| 9110-510, 9110-51A,9111-285, 9111-520, 9115-505|
9123-710, 9131-52A, 9405-520, 9406-520, 9406-525,
|8234-EMA, 9117-MMA, 9406-MMA|
|7998-60X, 7998-61X, 7778-23X|
|Power 7 Rack Servers||9119-FHB|
|AH730, AH760, AH780|
AL730, AL740, AL770
|8248-L4T, 8408-E8D, 8412-EAD|
|8202-E4C, 8231-E1C, 8246-L1C, 8246-L1S|
|8202-E4D, 8231-E1D, 8246-L1D, 8246-L1T|
|8205-E6B, 8231-E2B, 8246-L2B, 8268-E1D|
|8205-E6C, 8231-E2C, 8246-L2C, 8246-L2S|
|8205-E6D, 8231-E2D, 8246-L2D, 8246-L2T|
|Power 7 Blades||8406-70Y|
|Power 7 Flex Nodes||7895-23A|
This space evaluation includes a buffer to allow for about a 20% growth in the profile usage, which can be used for more partitions, more profiles or various other configuration changes.
If any server does not pass the space evaluation then the HMC should not be upgraded. Options include staying at HMC V7R7.7.0 (or upgrading no further then V7R7.7.0 for HMCs on earlier releases). Another alternative, is to free up space prior to the upgrade by deleting unused profiles and partitions.
For further information please contact your IBM service representative.
17 June 2018