hbase-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "stack (JIRA)" <j...@apache.org>
Subject [jira] Commented: (HBASE-1058) Prevent runaway compactions
Date Mon, 13 Apr 2009 19:27:14 GMT

    [ https://issues.apache.org/jira/browse/HBASE-1058?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12698510#action_12698510

stack commented on HBASE-1058:

So, this patch schedules a compaction if > threshold files at flush time.  It will not
flush until we are below threshold?  So it holds the flush up in memory until compaction completes?
 This makes it more likely the 'block client writes' gate comes down?    It makes it so our
blocking writes gate now comes down not only if we are > flush size times 2 (default) but
also if we are > number of files are > threshold?

Just trying to understand. 

J-D and Andrew, could it be that this patch will make us more stable at the cost of slowing
update rate?

> Prevent runaway compactions
> ---------------------------
>                 Key: HBASE-1058
>                 URL: https://issues.apache.org/jira/browse/HBASE-1058
>             Project: Hadoop HBase
>          Issue Type: Bug
>            Reporter: stack
>            Assignee: Andrew Purtell
>            Priority: Blocker
>             Fix For: 0.20.0
>         Attachments: hbase-1058-v2.patch, hbase-1058-v3.patch, hbase-1058.patch
> A rabid upload will easily outrun our compaction ability dropping flushes faster than
we can compact them up.  Fix.

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

View raw message