Not moved a single package on secondary node

[Goriik]

I have two nodes Nod_1 and Nod_2. Uploading two packages Pkg1 and Pkg2 on Nod_1. Pkg1 and Pkg2 are running on Nod_1. I make shutdown-r to Nod_1. Pkg1 perezzhaet on Nod_2, and Pkg1 goes into a halting and not moving for Nod_2. If you run the packages on Nod_2 and reboot Nod_2, then Pkg1 moves, and Pkg2 goes into a halting. How do you think what could be the problem? What should I check?

[Turgay Cavdar]

First you should check package log file, which resides in either /var/adm/cmcluster/log or in package direcrotr /etc/cmcluster/PKG*

[Izhal]

Check your /etc/cmcluster/pkg.... and
/var/adm/syslog/syslog.log.

[Rita C Workman]

This thread seems to be a somewhat repeat of an earlier thread:

https://forums11.itrc.hp.com/service/forums/questionanswer.do?threadId=1412081

...and unfortunately there you didn't post and cluster/pkg logfiles either...

Nobody can really help you Goriik until we get enough information to make reasonable suggestions.
Please post your logs .. and let's just do this on one thread.

Regards,
Rita

[Deeos]

could you check the following task over the server

cmviewcl -v

once halt the pkg1 on node1 and then run pkg1 on node1, wheather the package will run on node1 or not.

and tell me if you are using virtual Ip address and how it is defined

Is there any valid entry for virtual ip address?

[Deeos]

is there any entry in /etc/rc.config.d/netconf for virtual IP address?

[Kapil Jha]

what are the logs various logs says I am sure it would at least tell you wheather it tried to start on other node or not.
You may have to check below logs
1)pkg logs /etc/cmcluster/pkg_name/somethin.log
on both the server
2) check syslog for the same time.

and
what does cmviewcl says is AUTO_RUN enabled everytime.

BR,
Kapil+

[Goriik]

I do not understand what was the problem, but earned. Thank you all.

[Stephen Doud]

Package failover failures have the following general causes:

On the primary node:
1) Package shutdown on the primary node failed to complete normally. Look at the package control log on that node.

On the adoptive node:
2) The adoptive node was down, not running Serviceguard, or disabled from
receiving all packages or specific packages (see 4).
See the primary node's syslog.log for clues.

3) A package resource did not meet UP criteria on the adoptive node
- a package resource did not meet start criteria
- a package-dependent subnet was down
See the syslog.log for clues.

4) The package failover was not "enabled"
- the AUTO_RUN or Node_switching flag is/was disabled
See the syslog.log for an indication.
Example:
cmcld: Package cannot run on this node because switching has been disabled for this node


5) Package startup began on the adoptive node, but encountered a problem which forced it to halt. Check the package log for the reason.

To properly troubleshoot each possibility, identify the date and time when the
package failover failure occured (check the /var/adm/syslog/syslog.log and perhaps even OLDsyslog.log on the both nodes) and the package control log
(/etc/cmcluster/*/*.log or /var/adm/cmcluster/*log) on both the primary and adoptive nodes to get the full picture.