Operating System - HP-UX
1838703 Members
3414 Online
110128 Solutions
New Discussion

Packages automatically startup

 
Wayne Fontaine
Advisor

Packages automatically startup

While testing a new installation of a 4 node cluster, we noticed that all packages which were in a state of being down, halted and disowned automatically started up after I rebooted one of the nodes. Does anybody have an explanation for this?

Thanks,
Jim
4 REPLIES 4
Rita C Workman
Honored Contributor

Re: Packages automatically startup

Ah...but did you check and ensure via cmviewcl that AUTO_RUN was 'disabled' ?

So if AUTO_RUN in your ascii file is set to YES, and your cmviewcl showed that it was 'enabled' - Then yes it would automatically start those pkgs up after a reboot.

Rgrds,
Rita
Ermin Borovac
Honored Contributor

Re: Packages automatically startup

Check FAILBACK_POLICY in package configuration file. If set to AUTOMATIC SG will restart package on its primary node when primary node becomes capable on running the package.
Geoff Wild
Honored Contributor

Re: Packages automatically startup

From:

http://docs.hp.com/en/B3936-90100/index.html

Automatic cluster startup will take place if the flag AUTOSTART_CMCLD is set to 1 in the /etc/rc.config.d/cmcluster file. When any node reboots with this parameter set to 1, it will rejoin an existing cluster, or if none exists it will attempt to form a new cluster.

IMHO - it is a good thing that a node stays down if it reboots for any reason - why? you need to determine the cause of the reboot just in case something was wrong - which could cause issues in your packages.

Rgds...Geoff
Proverbs 3:5,6 Trust in the Lord with all your heart and lean not on your own understanding; in all your ways acknowledge him, and he will make all your paths straight.
Wayne Fontaine
Advisor

Re: Packages automatically startup

FAILBACK_POLICY is MANUAL on all the packages so none should have started up once the node became available.

The other parameters suggested to check is identical to another cluster which does not have the problem.

Unfortunately, the cluster is in full production now so I don't have the luxury of playing around with it.

Thanks for all the responses.

Regards,
Jim