[jira] [Updated] (IGNITE-4879) System pool starvation while partition evicting.

["Andrew Mashenkov (JIRA)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/IGNITE-4879?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Andrew Mashenkov updated IGNITE-4879:
-------------------------------------
    Attachment: thread-dump-example.log

> System pool starvation while partition evicting.
> ------------------------------------------------
>
>                 Key: IGNITE-4879
>                 URL: https://issues.apache.org/jira/browse/IGNITE-4879
>             Project: Ignite
>          Issue Type: Bug
>          Components: cache
>    Affects Versions: 1.7
>            Reporter: Andrew Mashenkov
>            Assignee: Andrew Mashenkov
>            Priority: Critical
>             Fix For: 2.0
>
>         Attachments: thread-dump-example.log
>
>
> Some system pool threads is waiting for others made partition eviction.
> We should change partition eviction in async way, to allow only one thread do useful job without blocking other threads.
> Startpoint is clearAll() call in GridDhtLocalPartition.tryEvict() method.
> TryEvict method is called synchronously during few basic operations that can stuck unexpectedly.



--
This message was sent by Atlassian JIRA
(v6.3.15#6346)