You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@ozone.apache.org by "Ritesh Shukla (Jira)" <ji...@apache.org> on 2023/03/30 05:05:00 UTC
[jira] [Commented] (HDDS-7987) Recon always fails to start because of memory overflow
[ https://issues.apache.org/jira/browse/HDDS-7987?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17706692#comment-17706692 ]
Ritesh Shukla commented on HDDS-7987:
-------------------------------------
Have you tried setting the memory to be higher? For a billion objects, recon might need more memory than 31 GB. Also, do you have a heap dump that you can share somehow?
> Recon always fails to start because of memory overflow
> ------------------------------------------------------
>
> Key: HDDS-7987
> URL: https://issues.apache.org/jira/browse/HDDS-7987
> Project: Apache Ozone
> Issue Type: Bug
> Components: Ozone Recon
> Affects Versions: 1.3.0
> Reporter: Max Xie
> Assignee: Devesh Kumar Singh
> Priority: Major
> Attachments: image-2023-02-17-10-46-11-333.png
>
>
> We deployed an ozone cluster with 10 dn nodes, conducted a stress test, and generated 1 billion keys under 1 bucket, which caused Recon to always overflow memory and the service was unavailable.
>
> Recon JVM monitor
> !image-2023-02-17-10-46-11-333.png!
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@ozone.apache.org
For additional commands, e-mail: issues-help@ozone.apache.org