You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@ignite.apache.org by "Moti Nisenson-Ken (Jira)" <ji...@apache.org> on 2019/08/28 14:47:00 UTC
[jira] [Commented] (IGNITE-8830) Non-coordinator node unable to
finish local exchange should detect it and stop
[ https://issues.apache.org/jira/browse/IGNITE-8830?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16917839#comment-16917839 ]
Moti Nisenson-Ken commented on IGNITE-8830:
-------------------------------------------
IGNITE-8828 is not implemented - is this actually a duplicate? Anything that can help prevent or mitigate PME resulting in cluster hang is important.
> Non-coordinator node unable to finish local exchange should detect it and stop
> ------------------------------------------------------------------------------
>
> Key: IGNITE-8830
> URL: https://issues.apache.org/jira/browse/IGNITE-8830
> Project: Ignite
> Issue Type: Improvement
> Components: general
> Reporter: Sergey Chugunov
> Priority: Major
> Labels: iep-25
> Original Estimate: 336h
> Remaining Estimate: 336h
>
> At the final stage of Partition Map Exchange coordinator node sends full partitions map to all other nodes.
> If some node fails to apply this message and finish its local exchange it won't be able to operate correctly.
> To prevent this node should be able to check status of this exchange on coordinator (by sending some diagnostic message). If the exchange has already finished on coordinator, node should stop.
--
This message was sent by Atlassian Jira
(v8.3.2#803003)