package is down in node1

[nishant_4]

hi,
i am well aware of service guard in hp-ux and have bit exposure of it but i don't know what step should be followed if i get issue that "package is down in node1".
please guide me with the necessary steps.

Thanks

[Ivan Krastev]

Check the reason for package down - see log file in /etc/cmcluster//package.sh.log file.

You can bring back package with 'cmrunpkg pkgname'.

See more for packages and cluster status - http://docs.hp.com/en/B9903-90054/ch07s01.html


regards,
ivan

[Todd McDaniel_1]

You should really go check the package log file under /etc/cmcluster//*log, just in case there was something wrong before you attempt to start the package again.That will show you what if any reason for the package failing.

Also, check your syslog to see if there were any hardware issues.

[Geoff Wild]

The package Log file is your friend.

Maybe the package is supposed to be down - used only for DR?

what do you see:

cmviewcl -v

Rgds...Geoff

[Shardha]

Dear Nishant,

Use
cmviewcl -v command to check the status of the package.
also use cmmodpkg -e pkgname to start and enable the package on a particular node.

Package down means, the vg is not enabled, files system is not mounted, IP is not assignied and if the application start up is configured, then application will not start.

Sahrdha

[whiteknight]

nishant,

the most important logs to check

1)/var/adm/syslog/syslog.log
2) The package control logs are usually at
/etc/cmcluster/*/*.log
3) cmviewcl -v

4) serviceguard recovery handbook
http://www1.itrc.hp.com/service/cki/docDisplay.do?docLocale=en&docId=emr_na-c01037931-1

WK
please assign points

[nishant_4]

Hi all,

thank you for your replies. i have already comfigure clustering in test servers but i don't know what process all you guys follow if u get this issue. Actually i am looking for the sequence of step which u follow generally so that i couldn't make mistake in productio servers.