Galder Zamarreño <galder.zamarreno> made a comment on jira ISPN-2752

A way to silence this scenario is needed, which should be included as part of the fix for ISPN-2577.

Galder Zamarreño <galder.zamarreno> made a comment on jira ISPN-2752

IRC discussion: https://gist.github.com/647ec76c9e3d7cc28ef3

Mircea Markus <mmarkus> made a comment on jira ISPN-2752

[~mlinhard] [~dan.berindei] this exception is received on a node that is broadcasting a TopologyControllCommand (e.g. consistent hash update, rebalance start). By the look of code this is sent synchronously and should impact the command execution. Does't this affect the rebalancing in any way?

I haven't seen this in any scenario where it would affect data integrity. I considered this low prio/severity cause it only happened on cluster shutdown.

Dan Berindei <dberinde> made a comment on jira ISPN-2752

[~mircea.markus], the exception message "Remote (node0005/default) failed unexpectedly" is misleading. In fact the coordinator is shutting down shortly after it became coordinator, and while an asynchronous listener thread is trying to retrieve the status in the cluster, the JChannel was closed and it also closed all pending (outgoing) requests with an IllegalStateException.

Since the exception is local to the coordinator, which is already shutting down, this is not going to impact anything. We could probably add a check when catching this exception and skip logging it if the cache manager is shutting down.

Set flag to nominate this bug for 6.2 release notes.