TSAMP + DB2 HADR : What happens when Network Tie-breaker and netmon.cfg file IPs are not reachable.
Environment Details: Two node (AIX and Linux) DB2 HADR v10.5 * TSAMP v3.2.2 and v4.1
Description: TSAMP is configured to handle automation of two node HADR setup. Network tie-breaker is defined. Tie-breaker IP is a third IP outside the two node network configuration. I.e Network switch IP address * netmon.cfg configured to have same IP as Tie-breaker.
Now when there is a network problem, to resolve the quorum, if TSAMP is unable to reach tie-breaker and also none of the IP in netmon.cfg what happens in this case.
Also, is it possible to have a two different type of tie-breaker for a given cluster. i.e 1) Network tie-breaker 2) Disk tie-breaker.
Thanks
Although you can define several tiebreaker resources in the resource class IBM.TieBreaker, only one of them can be active in the cluster at the same time.
Tivoli System Automation for Multiplatforms 4.1.0 - Configuring - Configuring the tiebreaker
Thank you Clark for getting back.
One other point, is it possible to have two tie-breaker for single cluster of different type.
Thanks
If the network tiebreaker and the other node are unreachable then the node will be rebooted due to critical resource protection.
You will see messages in syslogs with the switch to "NO_QUORUM" and since all DB2 related resources are created as Critical Resources (ProtectionMode=1)( lsrsrc -Ab IBM.Application Name ProtectionMode ) then any online resource will result in a reboot should the node lose quorum.
The netmon.cf files are principally used to prove that a local NIC is functioning and really is not going to impact quorum determination at all, so regardless of whether the IPs in the netmon.cf are pingable or not the node that loses quorum is going to be rebooted.