You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@ignite.apache.org by "Pavel Voronkin (JIRA)" <ji...@apache.org> on 2018/10/22 13:12:00 UTC
[jira] [Commented] (IGNITE-9870)
GridDhtPartitionsFullMessage#prepareMarshal compression parallelization.
[ https://issues.apache.org/jira/browse/IGNITE-9870?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16658970#comment-16658970 ]
Pavel Voronkin commented on IGNITE-9870:
----------------------------------------
We can parallelize each marshall and zip of "parts", "partCntrs", "partCntrs2", "partHistSuppliers" and "partsToReload" in a separate task in thread pool and speed up message preparation.
> GridDhtPartitionsFullMessage#prepareMarshal compression parallelization.
> ------------------------------------------------------------------------
>
> Key: IGNITE-9870
> URL: https://issues.apache.org/jira/browse/IGNITE-9870
> Project: Ignite
> Issue Type: Improvement
> Components: general
> Affects Versions: 2.6
> Reporter: Stanilovsky Evgeny
> Assignee: Stanilovsky Evgeny
> Priority: Major
>
> In huge topologies ~ 200 cluster nodes , 100 caches , 32k partitions per cache, full map generation takes about 3 sec, seems correct approach here in parallelization.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)