concurrency violation

[allaboutunix]

Hi Team,

This alert came in my environment in one of the AIX server 6.1 of concurrency violation.

Subject: VCS SevereError for Service Group sapgtsprd, Service group concurrency violation

Event Time: Wed Dec  3 06:46:54 EST 2014 Entity Name: sapgtsprd Entity Type: Service Group Entity Subtype: Failover Entity State: Service group concurrency violation Traps Origin: Veritas_Cluster_Server System Name: mapibm625 Entities Container Name: GTS_Prod Entities Container Type: VCS

==================================================================================

engineA.log are,

2014/12/03 06:46:54 VCS INFO V-16-1-10299 Resource App_saposcol (Owner: Unspecified, Group: sapgtsprd) is online on mapibm625 (Not initiated by VCS)
2014/12/03 06:46:54 VCS ERROR V-16-1-10214 Concurrency Violation:CurrentCount increased above 1 for failover group sapgtsprd
2014/12/03 06:46:54 VCS NOTICE V-16-1-10233 Clearing Restart attribute for group sapgtsprd on all nodes
2014/12/03 06:46:55 VCS WARNING V-16-6-15034 (mapibm625) violation:-Offlining group sapgtsprd on system mapibm625
2014/12/03 06:46:55 VCS INFO V-16-1-50135 User root fired command: hagrp -offline sapgtsprd  mapibm625  from localhost
2014/12/03 06:46:55 VCS NOTICE V-16-1-10167 Initiating manual offline of group sapgtsprd on system mapibm625

======================================================================================================

What is Concurrency Violation in VCS? What are the steps we should have to take to resolve this?Kindly explain in detail.

Thanks,

Allaboutunix

[mikebounds · Accepted solution]

Most service groups are of type failover (the default) meaning it should only run on one system at any one time and it can fail over to another system, so if VCS detects a failover group is running on more than one node then it reports a Concurrency Violation and then tries to offline on the node that VCS did not online it on.

So in your situation resource App_saposcol in group sapgtsprd was onlined on mapibm625 outside of VCS control when it was already online on another system in the cluster so VCS offlined it:

2014/12/03 06:46:55 VCS NOTICE V-16-1-10167 Initiating manual offline of group sapgtsprd on system mapibm625

If group should run on 2 systems at the same time then you should configure group as Parallel (set group attribute Parallel = 1) and then group will run on all systems in the cluster at the same time.

Mike

[Gaurav_S · Accepted solution]

Mike has perfectly pointed the cause , To add to what Mike said, if application has to run on more than 1 node, you need to enable parallel attribute. However the application should have intelligence to run on multiple nodes at the same time without causing issues to data because multiple application instances may access same database underneath which may cause issues to DB,

G

[allaboutunix]

then tries to offline on the node that VCS did not online it on >>>>>>>>

so it means if the state of Concurrency Violation arrived  then vcs will auto-correct this issue, if there will be failover service group.Is it or correct me if I am wrong?

Thanks

Allaboutunix

[mikebounds]

Correct - vcs will auto-correct Concurrency Violation by offlining resource on the node that it came online outside of VCS.

Mike