–# V7R7.2 to V7R7.6, or from V7 to V8 for example.
You can find network install resources right here :
…or there :
Backup your data !!!
# bkconsdata -r ftp -h [ftp host] -u [user] -d [directory]
As said on this blog , this is a good practice :
These commands will cleanup the events and the temporary files in order to make some space.
Then you can reboot before upgrading (not mandatory I think)
# chsvcevent -o closeall # chhmcfs -o f -d 0 # hmcshutdown -t now -r
Check the need of a 2-step migration
Sometimes you might need to upgrade to an intermediate HMC level, before reaching to the desired level :
For example, V7R7.9.0 requires a minimal level of V7R7.6.0 . If your HMC is V7R6.xx, you need to upgrade firstly to V7R7.6.0, before going to V7R7.9.
Check if there is enough space
Upgrade commands on HMC are not the verbose type, so you have to be sure that you have sufficient space on your hard drive (sometimes on old configuration it is very small).
1. Prior to upgrading from V7R7.6.0 or V7R7.7.0, users with many partitions and partition profiles must verify that their current server configuration is compatibile with HMC Version 7 Release 7.8.0 and later. The HMC uses a storage area on each managed server to persist the server configuration. HMC V7R7.8.0 and later increases the usage of this storage area. Servers already containing a large number of profiles may not have sufficient space to allow an upgrade to HMC Version 7 Release 7.8.0 and later.
New command to check upgrade sizing : lsprofspace (only starting with V7R7.6)
please see HMC Version 7 Release 7.8.0 Upgrade sizing
Better shut down the other HMC (I never do that, but, you know… 🙂 )
When two HMCs manage the same server, both HMCs must be at the same version. Once the server is connected to the higher version of the management console, the partition configuration is migrated to the latest version. Lower management consoles will not be able to understand the data properly and may fail. Before upgrading an HMC, you must disconnect the redundant HMC from the managed systems. Once the first upgrade is complete, you can upgrade the disconnected HMC, then reconnect it to the managed systems when the HMCs are at the same level again.
Check Firmware compatibility with Target HMC level
In order to know the various managed system’s firmwares that belong to your HMC, please refer to my post on HMC CLI cheatsheet , but I’ll put it back here for you :
# lssyscfg -r sys -Fname |while read f do model=`lssyscfg -r sys -Fname:type_model |grep $f` echo -n "$model: ";lslic -m $f -t sys -Fcurr_ecnumber_primary:activated_level done frame,9117-MMC: 01AM740:100 frame,9117-MMD: 01AM760:51 frame,9117-MMD: 01AM760:68 frame,8202-E4C: 01AL740:152 frame,9117-MMD: 01AM760:68 frame,9117-MMD: 01AM760:68 frame,9117-MMD: 01AM760:51 frame,8231-E1D: 01AL770:90 frame,8202-E4D: 01AL770:90 frame,8408-E8D: 01AM770:48 frame,8202-E4C: 01AL740:152 frame,8231-E1D: 01AL770:48 frame,8231-E1D: 01AL770:90 frame,9117-MMD: 01AM760:68 frame,8231-E1D: 01AL770:90 frame,8202-E4D: 01AL770:90 frame,8231-E1D: 01AL770:90 frame,9117-MMC: 01AM740:100
Fix Level Recommendation Tool (FLRT)
Please check FLRT site to get the right levels regarding which firmware you’ve got :
You can also check the links section for the power matrix code for each generation of power , but these are static pages, you better
use FLRT now.
Let’s get it done
Once the files are uploaded to a NIM server for example , you can use getupgfiles (add the -s option to use ftps instead of ftp):
hscroot@hmc:~#getupgfiles -h [FTP server] -u [ftpuser] -s -d /HMC_Update_V8R810_SP1/network_install Enter the current password for user [ftpuser]: hscroot@hmc:~#
Meanwhile you can check the progress of this copy by opening another terminal and typing this :
hscroot@hmc:~#while true ; do date; ls -la /hmcdump /hmcboot;monhmc -r disk -n 0; sleep 10; done
This operation will mount /mnt/upgrade and save configuration data then unmount /mnt/upgrade. This should take ony a few moments.
(This data will be used during next boot, so do not initiate reboot after this has been done, or you have to do this step again.)
hscroot@hmc:~#saveupgdata -r disk hscroot@hmc:~#
Give me some space
ATTENTION !!! If at this step, you get this kind of error :
hscroot@hmc:~#saveupgdata -r disk The following error occured: The Save Upgrade Data request failed.
hscroot@hmc saveupgdata -r disk The following error occured: There is no more space left on the upgrade partition. Please remove files saved in users' home directory and retry the operation.</pre>
!!!! You shall not reboot the HMC , stop the procedure right there .
Otherwise the HMC won’t come back at all. It will be like a factory reset, with
all default options (no SSH, no authentication, no more user profiles, no more DHCP server…)
So you have to clean up the filesystems on the HMC. if you can’t, then , you’re likely to keep your
actual HMC level.
All cleaned up ? let’s go on
if the saveupgdata -r disk ended up OK, you can follow these steps now:
hscroot@hmc:~#chhmc -c altdiskboot -s enable --mode upgrade hscroot@hmc:~# hmcshutdown -r -t now hscroot@hmc:~# Broadcast message from root@hmc (Fri Sep 12 14:57:41 2014): The system is shutting down for reboot now.
This is the freaky part, because from now on and for a bunch of minutes (likely 20-30 minutes), you go blind.
No ping, no terminal, nothing to tell you whether the upgrade is ongoing, or if the HMC is completely crashed.
Keep the faith, because after 20-30 minutes, you will get back your ssh access.
Check the version
hscroot@hmc:~#lshmc -v "vpd=*FC ???????? *VC 20.0 *N2 Fri Sep 12 14:44:08 CEST 2014 *FC ???????? *DS Hardware Management Console *TM 7042-CR8 *SE 99E199C *MN IBM *PN 434097 *SZ 8399999994 *OS Embedded Operating Systems *NA 100.100.100.100 *FC ???????? *DS Platform Firmware *RM V8R220.127.116.11 " hscroot@hmc:~#lshmc -V "version= Version: 8 Release: 8.1.0 Service Pack: 0 HMC Build level 20140602.3 MH01421: Required fix for HMC V8R8.1.0 (06-03-2014) ","base_version=V8R8.1.0 "
–> service pack, ou security/mandatory fix installation
if you have a FTP server up :
updhmc -t s -i -h [FTPserver]-u [ftpuser] -f /HMC_Update_V8R810_SP1/HMC_Update_V8R810_SP1.iso
No FTP server, or FTPS (because updhmc can’t do ftps, unlike getupgfiles, don’t ask me why):
–> you have to copy locally the .iso files :
user@NIM:/export/nim/HMC_Update_V8R810_SP1 sudo scp HMC_Update_V8R810_SP1.iso \ hscroot@hmc:/home/hscroot HMC_Update_V8R810_SP1.iso hscroot@hmc:~#updhmc -t l -f /home/hscroot/HMC_Update_V8R810_SP1.iso iptables: Chain already exists. ip6tables: Chain already exists. iptables: Chain already exists. ip6tables: Chain already exists. iptables: Chain already exists. ip6tables: Chain already exists. iptables: Chain already exists. ip6tables: Chain already exists. The corrective service file was successfully applied. A mandatory reboot is required but was not specified on the command syntax. hscroot@hmc:~#hmcshutdown -t now -r
Vérifier avec lshmc -v.
Power Matrix code (still useful but deprecated):