You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@ignite.apache.org by "Ignite TC Bot (Jira)" <ji...@apache.org> on 2021/02/07 12:52:00 UTC

[jira] [Commented] (IGNITE-14131) IgniteCompute tasks with same name, running from one node and different ClassLoaders can lead to OOM. Fix problems with concurrent ignite.compute call.

    [ https://issues.apache.org/jira/browse/IGNITE-14131?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17280491#comment-17280491 ] 

Ignite TC Bot commented on IGNITE-14131:
----------------------------------------

{panel:title=Branch: [pull/8759/head] Base: [master] : Possible Blockers (1)|borderStyle=dashed|borderColor=#ccc|titleBGColor=#F7D6C1}
{color:#d04437}Cache 6{color} [[tests 1|https://ci.ignite.apache.org/viewLog.html?buildId=5858543]]
* IgniteCacheTestSuite6: CacheExchangeMergeTest.testDelayExchangeMessages - Test has low fail rate in base branch 0,0% and is not flaky

{panel}
{panel:title=Branch: [pull/8759/head] Base: [master] : No new tests found!|borderStyle=dashed|borderColor=#ccc|titleBGColor=#F7D6C1}{panel}
[TeamCity *--&gt; Run :: All* Results|https://ci.ignite.apache.org/viewLog.html?buildId=5858284&amp;buildTypeId=IgniteTests24Java8_RunAll]

> IgniteCompute tasks with same name, running from one node and different ClassLoaders can lead to OOM. Fix problems with concurrent ignite.compute call.
> -------------------------------------------------------------------------------------------------------------------------------------------------------
>
>                 Key: IGNITE-14131
>                 URL: https://issues.apache.org/jira/browse/IGNITE-14131
>             Project: Ignite
>          Issue Type: Improvement
>          Components: compute
>    Affects Versions: 2.9.1
>            Reporter: Stanilovsky Evgeny
>            Assignee: Stanilovsky Evgeny
>            Priority: Major
>          Time Spent: 10m
>  Remaining Estimate: 0h
>
> The root cause of such a problem is growing from assumption that one node can obtain only one class loader per class name. Thus multiple tasks calling with different classloaders leads to huge cache grow in server side and finally leads to oom with jvm metaspace. Additionally we can`t use p2p from multiple threads, for example ignite instance is shared as a spring bean.
>  



--
This message was sent by Atlassian Jira
(v8.3.4#803005)