cancel
Showing results for 
Search instead for 
Did you mean: 
Read only

Real world experience using FAULT TOLERANCE from VMWARE for the ASCS instance

mamartins
Active Contributor
0 Likes
136

Currently, system is installed on 2 node physical cluster with HPE Service Guard cluster. ASCS instance is a cluster resource and each node have an ERS instance to replicate the enqueue to the other node.

New infrastructure is virtual on top of VMWARE, and instead of replicating the cluster, in order to simplify the solution, FT was implemented on the VM running the ASCS instance. There is another VM just for the DB and 3 VM for PAS and AAS.  During the testing phase (VMOTION's and FT simulation) we got mixed results. Initially, everything seems OK, but after some time, new scheduled jobs got stuck waiting for the ENQUEUE, like this:

mamartins_1-1767950250610.png

To recover from this, a restart of the system is needed. During the FT operations, we didn't saw any packet drops, but obviously the PING took longer than usual, from 5 ms to 150ms (only one). 

 

Accepted Solutions (0)

Answers (0)