[
https://issues.apache.org/jira/browse/CASSANDRA-5494?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13641596#comment-13641596
]
Ondřej Černoš commented on CASSANDRA-5494:
------------------------------------------
If only those where drops were detected are reported, why are there the zeros?
> Inconsistent view of nodes in CommandDroppedTasks
> -------------------------------------------------
>
> Key: CASSANDRA-5494
> URL: https://issues.apache.org/jira/browse/CASSANDRA-5494
> Project: Cassandra
> Issue Type: Bug
> Components: Core
> Affects Versions: 1.2.3
> Reporter: Ondřej Černoš
>
> Please see related CASSANDRA-5493 for context.
> On our testing platform we saw output from {{get CommandDroppedTasks}} in
> jmxterm, which was different on different DCs. In DC1, we saw 3x public IPs
> of remote DC nodes, 2x public IPs of local DC nodes, 2x private IPs of local
> DC nodes. We don't see stats for the very node the command is run on.
> On the other hand, on the remote DC, we see stats for the node the command is
> run on, stored under the node's public IP.
> This is at least confusing, maybe even a bug.
> After restarting all the nodes in the cluster (both datacenters) the
> situation settled - now we don't see stats for the node the command is run
> on. This is however inconsistent with the situation on production (see
> CASSANDRA-5493).
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
