Hi This fixes a bug that Andrew Beekhof found when using pacemaker.
The situation is like this: 4 nodes all up and joined a cpg group. in the cpg config handler you try to send messages you shut them all down at the same time Sometimes you get a case where 1 or 2 nodes' corosync won't shutdown and all API respond with TRY_AGAIN. The reason for this happening is corosync gets stuck in a sync state depending on the timing of the sync abort. patches to follow. Regards -Angus _______________________________________________ Openais mailing list [email protected] https://lists.linux-foundation.org/mailman/listinfo/openais
