hbase-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Jean-Daniel Cryans (JIRA)" <j...@apache.org>
Subject [jira] Commented: (HBASE-1058) Prevent runaway compactions
Date Tue, 14 Apr 2009 23:52:14 GMT

    [ https://issues.apache.org/jira/browse/HBASE-1058?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12698987#action_12698987
] 

Jean-Daniel Cryans commented on HBASE-1058:
-------------------------------------------

No problem.

What I mean is that we should jump start that thread to a defined value.  For example:

1. While flushing, we see that we have > than 4 stores files so we ask for a compaction
and wait for it. The compaction of 4 files took 2 minutes.

2. While flushing, we see that we have > than 4 stores files so we start the CompactionLimitThread
and set it to start on cell 5 of the limitSteps array. We then ask for a compaction and wait
for it. The compaction of 2 files took less than 2 minutes.

Currently we are at 1, I think of something like 2.

> 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-v4.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.


Mime
View raw message