You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@cassandra.apache.org by "Jonathan Ellis (JIRA)" <ji...@apache.org> on 2009/12/08 17:29:18 UTC

[jira] Commented: (CASSANDRA-609) Easy to OOM on log replay since memtable limits are ignored

    [ https://issues.apache.org/jira/browse/CASSANDRA-609?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12787562#action_12787562 ] 

Jonathan Ellis commented on CASSANDRA-609:
------------------------------------------

To clarify: under low load, you won't much exceed a single memtable's worth of inserts since when a memtable flushes it marks the commitlog header as "start replay from here."  But under high load, you can have a significant amount of inserts done while flushes are queued up in sort + write executors, and this is where you run into trouble on replay.

> Easy to OOM on log replay since memtable limits are ignored
> -----------------------------------------------------------
>
>                 Key: CASSANDRA-609
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-609
>             Project: Cassandra
>          Issue Type: Bug
>          Components: Core
>            Reporter: Jonathan Ellis
>            Assignee: Jonathan Ellis
>             Fix For: 0.5
>
>         Attachments: 609.patch
>
>


-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.