You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@ignite.apache.org by "Sergey Chugunov (JIRA)" <ji...@apache.org> on 2018/04/24 09:36:00 UTC

[jira] [Comment Edited] (IGNITE-8372) Cluster metrics are reported incorrectly on joining node with ZK-based discovery

    [ https://issues.apache.org/jira/browse/IGNITE-8372?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16449530#comment-16449530 ] 

Sergey Chugunov edited comment on IGNITE-8372 at 4/24/18 9:35 AM:
------------------------------------------------------------------

Patch artifacts: [PR|https://github.com/apache/ignite/pull/3907], [upsource review|https://reviews.ignite.apache.org/ignite/review/IGNT-CR-585], [TC run|https://ci.ignite.apache.org/viewQueued.html?itemId=1240990&tab=queuedBuildOverviewTab]


was (Author: sergey-chugunov):
Patch artifacts: [PR|https://github.com/apache/ignite/pull/3907], [upsource review|https://reviews.ignite.apache.org/ignite/review/IGNT-CR-585] [TC run|https://ci.ignite.apache.org/viewQueued.html?itemId=1240990&tab=queuedBuildOverviewTab]

> Cluster metrics are reported incorrectly on joining node with ZK-based discovery
> --------------------------------------------------------------------------------
>
>                 Key: IGNITE-8372
>                 URL: https://issues.apache.org/jira/browse/IGNITE-8372
>             Project: Ignite
>          Issue Type: Bug
>          Components: zookeeper
>            Reporter: Sergey Chugunov
>            Assignee: Sergey Chugunov
>            Priority: Minor
>             Fix For: 2.5
>
>
> When new node joins with ZK discovery it sometimes reports negative number of CPUs and incorrect heap size.
> Message in log looks like this:
> {noformat}
> [myid:] - INFO  [disco-event-worker-#61:Log4JLogger@495] - Topology snapshot [ver=100, servers=100, clients=0, CPUs=-6, heap=0.5GB]
> {noformat}
> There is a race though between this report and ClusterMetricsUpdateMessage: if the node receives and process this message first which happens in a separate thread correct values are printed to log.



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)