Originally posted by: alaix

Hello
I just try to do an update from hacmp 5.2 to 5.4.0 (first) then I will apply the fixpack 5.4.1.

I have stop the cluster and use smitty update_all to update HACMP (rsct already on 2.4.9 - AIX 5.3 TL8 (5300-08-03-0831) ).

I had most of them successfully updated (smitty update_all -> status ok) but then when I do lslpp cluster*
I have some of them not important still on 5.2 (I guess I can install those one manually on 5.4 if I can find it)
lslpp -l cluster*
Fileset Level State Description

---

Path: /usr/lib/objrepos
cluster.adt.es.client.include
5.4.0.0 COMMITTED ES Client Include Files
cluster.es.cfs.rte 5.4.0.0 COMMITTED ES Cluster File System Support
cluster.es.client.lib 5.4.0.0 COMMITTED ES Client Libraries
cluster.es.client.rte 5.4.0.0 COMMITTED ES Client Runtime
cluster.es.client.utils 5.4.0.0 COMMITTED ES Client Utilities
cluster.es.client.wsm 5.4.0.0 COMMITTED Web based Smit
cluster.es.cspoc.cmds 5.4.0.0 COMMITTED ES CSPOC Commands
cluster.es.cspoc.dsh 5.4.0.0 COMMITTED ES CSPOC dsh
cluster.es.cspoc.rte 5.4.0.0 COMMITTED ES CSPOC Runtime Commands
cluster.es.plugins.dhcp 5.4.0.0 COMMITTED ES Plugins - dhcp
cluster.es.plugins.dns 5.4.0.0 COMMITTED ES Plugins - Name Server
cluster.es.plugins.printserver
5.4.0.0 COMMITTED ES Plugins - Print Server
cluster.es.server.cfgast 5.4.0.0 COMMITTED ES Two-Node Configuration
Assistant
cluster.es.server.diag 5.4.0.0 COMMITTED ES Server Diags
cluster.es.server.events 5.4.0.0 COMMITTED ES Server Events
cluster.es.server.rte 5.4.0.0 COMMITTED ES Base Server Runtime
cluster.es.server.testtool
5.4.0.0 COMMITTED ES Cluster Test Tool
cluster.es.server.utils 5.4.0.0 COMMITTED ES Server Utilities
cluster.es.worksheets 5.4.0.0 COMMITTED Online Planning Worksheets
cluster.license 5.4.0.0 COMMITTED HACMP Electronic License
cluster.msg.En_US.cspoc 5.4.0.0 COMMITTED HACMP CSPOC Messages - U.S.
English IBM-850
cluster.msg.en_US.cspoc 5.2.0.0 COMMITTED HACMP CSPOC Messages - U.S.
English
cluster.msg.en_US.es.client
5.2.0.0 COMMITTED ES Client Messages - U.S.
English
cluster.msg.en_US.es.server
5.2.0.0 COMMITTED ES Recovery Driver Messages -
U.S. English

Path: /etc/objrepos
cluster.es.client.lib 5.4.0.0 COMMITTED ES Client Libraries
cluster.es.client.rte 5.4.0.0 COMMITTED ES Client Runtime
cluster.es.cspoc.rte 5.4.0.0 COMMITTED ES CSPOC Runtime Commands
cluster.es.server.diag 5.4.0.0 COMMITTED ES Server Diags
cluster.es.server.events 5.4.0.0 COMMITTED ES Server Events
cluster.es.server.rte 5.4.0.0 COMMITTED ES Base Server Runtime
cluster.es.server.utils 5.4.0.0 COMMITTED ES Server Utilities

Path: /usr/share/lib/objrepos
cluster.man.en_US.es.data 5.4.0.0 COMMITTED ES Man Pages - U.S. English

but the output of lppch is not good on some filesets appears BROKEN
lppchk -v
lppchk: The following filesets need to be installed or corrected to bring
the system to a consistent state:

cluster.es.cspoc.cmds 5.2.0.7 (BROKEN)
cluster.es.cspoc.dsh 5.2.0.2 (BROKEN)
cluster.es.cspoc.rte 5.2.0.0 (not installed; requisite fileset)
cluster.es.cspoc.rte 5.2.0.6 (BROKEN)
but If I check those filesets on the system:
lslpp -l cluster.es.cspoc*
Fileset Level State Description

---

Path: /usr/lib/objrepos
cluster.es.cspoc.cmds 5.4.0.0 COMMITTED ES CSPOC Commands
cluster.es.cspoc.dsh 5.4.0.0 COMMITTED ES CSPOC dsh
cluster.es.cspoc.rte 5.4.0.0 COMMITTED ES CSPOC Runtime Commands

Path: /etc/objrepos
cluster.es.cspoc.rte 5.4.0.0 COMMITTED ES CSPOC Runtime Commands

there are all on 5.4.0.0 ... Is this normal ???

I didn;t reboot yet the server.

I don;t know if this means it is successfull or not... I don;t want to reboot and start the update on the primary node prio to be sure it is fine.

Any ideas how to fix BROKEN status ?
Cheers
Al

Originally posted by: Casey_B

Hello Al,

Some other filesets that appear to be at 5.2 are the message catalogs. (cluster.msg.en_US.*)

HACMP migrations can be tricky. There are changes that the software needs to do with the
HACMp odm's, and the installed software.

I think it would be best if you open a PMR with IBM support. THey will be able to examine your smit logs
to determine if there were any install errors, and also look at your HA config to see how the migration
is progressing.

Any information you get over the forum for this problem would likely be incomplete.

Hope this helps,
Casey

Originally posted by: grukrz1

I also had problem but with upgrade 5.4.1.0 -> 5.4.1.5 - the hint in smit.log was:

BROKEN REQUISITES: The following filesets are requisites of one or more
of the selected filesets listed above. They are currently installed
on the system but are in the BROKEN state.
Alternative actions that can be used to correct broken filesets are:
1. If the broken fileset is an update, re-install the update with options
to the installation program that will apply, commit and automatically
install requisite software without saving replaced files: -acgN flags.
2. Re-install the entire fileset, i.e., the base level and any desired
updates (may need to use the "FORCE" option: -F flag).
3. Deinstall the entire fileset (using "Remove Software Products":
-u flag).

cluster.es.cspoc.cmds 5.4.1.5 # ES CSPOC Commands
In my case the problem was linked to cl_passwd utlitity...

I've got hint on a hacmp forum to rename one file and re-run again the installation of the the broken fileset cluster.es.cspoc.cmds - it completed OK and all cluster.* filesets became COMMITED (BROKEN disappeared):

The file which to be renamed before one more installation startup was: mv /usr/es/sbin/cluster/etc/clpasswd/usr_bin_passwd.orig /usr/es/sbin/cluster/etc/clpasswd/blee_usr_bin_passwd.orig

hope this hint will help you as well...

kind regards,
K.