Sponsored Content
Full Discussion: upgrade hacmp
Operating Systems AIX upgrade hacmp Post 302484792 by shockneck on Monday 3rd of January 2011 08:40:01 AM
Old 01-03-2011
Quote:
Originally Posted by zxmaus
[...]I need to upgrade one of my clusters from HACMP 5.41 to HACMP 6.1.
Is there an easy + clean way to do it ? Any hurdles I have to anticipate? What should I read - anything I found so far seems to talk only about HACMP 5.x - and looks pretty straight forward.[...]
As always the answer is likely to be "it depends" Smilie So just some thoughts in advance...

By going from 5.4 to 6.1 you upgrade the version. So there might be more to consider than if you would just upgrade the release. It is almost impossible to foresee what might happen at what stage. E.g. do you upgrade RSCT also and if yes, how does upgrading RSCT work out? So to really know what is going to happen this has to be tested in a testing environment. If you can test, a rolling migration might be your choice.

In case you cannot test the upgrade with clones I would not use any online migration method offered by IBM but recreate the cluster. This is not as much work as it sounds. You can use either snapshot or (if IBM did not change the XML files from 5.4 to 6.1 again) online planning worksheet. Both methods keep the cluster definition so after installing the necessary software packages recreation of the cluster itself should not last langer than about 60 seconds. From my personal experiences this always works as long as there are no configuration errors (logical glitches, reservation probs, violation of networking rules, changes with snmp to name a few). So I would create snapshot and OLPW file from cluster. Uninstall HA software, install HA Software. Reboot. Check storage and network. Recreate Cluster from (convertet) snapshot or OLPW whichever suits you more.

Also remember that you can use alt_disk_install to keep the current setup for easy/fast rollback. Needless to say that you do make a backup of every cluster node before you start.
This User Gave Thanks to shockneck For This Post:
 

8 More Discussions You Might Find Interesting

1. AIX

Upgrade Hacmp to 5.4

Anyone ever did the upgrade before from 5.1? Any tips? (0 Replies)
Discussion started by: adiriot
0 Replies

2. AIX

Hacmp

(0 Replies)
Discussion started by: mhenryj
0 Replies

3. AIX

hacmp upgrade with db2

Hi All, My company is going to upgrade the db2 on the one node hacmp, I think there's one backup. After the upgrade of the db2. Do I need to restart cluster and server? I mean, do a graceful shutdown of the cluster, reboot the server and do a cluster start. Is that it? Do I need to reboot... (1 Reply)
Discussion started by: itik
1 Replies

4. AIX

HACMP

Hi, Can we use network for heartbeat, I mean can we use different network card for heartbeat. (6 Replies)
Discussion started by: vjm
6 Replies

5. AIX

HACMP

hi can anyone explain the concepts of HACMP and configuration (step by step) (2 Replies)
Discussion started by: udtyuvaraj
2 Replies

6. AIX

MQ upgrade(ver.6to7) in a HACMP cluster

Hi What is the procedure to upgrade the MQ from 6 to 7 in aix hacmp cluster. Do i need to bring down the cluster services running in both the nodes and then give #smitty installp in both the nodes separately. Please assist... (0 Replies)
Discussion started by: samsungsamsung
0 Replies

7. AIX

HACMP does not start db2 after failover (db2nodes not getting modified by hacmp)

hi, when I do a failover, hacmp always starts db2 but recently it fails to start db2..noticed the issue is db2nodes.cfg is not modified by hacmp and is still showing primary node..manually changed the node name to secondary after which db2 started immediately..unable to figure out why hacmp is... (4 Replies)
Discussion started by: gkr747
4 Replies

8. AIX

HACMP 5.4.1->5.5 offline upgrade - different instance numbers ?

Hello, I did offline HACMP(PowerHA) upgrade 5.4.1 to 5.5 - basically stopped HACMP services and upgraded cluster.* filesets. Tried to start services again - topsvcs refused to start on second node complaining that node instance numbers are different - and indeed they are different. This is... (2 Replies)
Discussion started by: vilius
2 Replies
GROUPD(8)							      cluster								 GROUPD(8)

NAME
groupd - compatibility daemon for fenced, dlm_controld and gfs_controld SYNOPSIS
groupd [OPTIONS] DESCRIPTION
The groupd daemon and libgroup library are used by the fenced, dlm_controld and gfs_controld daemons when they are operating in clus- ter2-compatible mode to perform a rolling cluster upgrade from cluster2 to cluster3. See cman(5) for more information on the upgrading configuration option needed to perform a rolling upgrade. When the upgrading option is enabled, cman adds the following to the online configuration: <group groupd_compat="1"/> This setting causes the cman init script to start the groupd daemon, and causes the groupd, fenced, dlm_controld and gfs_controld daemons to operate in the old cluster2 mode so they will be compatible with cluster2 nodes in the cluster that have not yet been upgraded. The upgrading setting, including the groupd_compat setting, cannot be changed in a running cluster. The entire cluster must be taken off- line to change these because the new cluster3 default modes are not compatible with the old cluster2 modes. The upgrading/compat settings cause the new cluster3 daemons to run the old cluster2 code and protocols. OPTIONS
Command line options override a corresponding setting in cluster.conf. -D Enable debugging to stderr and don't fork. -L Enable debugging to log file. -g num groupd compatibility mode, 0 off, 1 on. Default 0. -h Print a help message describing available options, then exit. -V Print program version information, then exit. SEE ALSO
cman(5), fenced(8), dlm_controld(8), gfs_controld(8) cluster 2009-01-19 GROUPD(8)
All times are GMT -4. The time now is 10:16 PM.
Unix & Linux Forums Content Copyright 1993-2022. All Rights Reserved.
Privacy Policy