hbase-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "ramkrishna.s.vasudevan (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HBASE-17777) TestMemstoreLAB#testLABThreading runs too long for a small test
Date Fri, 26 May 2017 07:07:04 GMT

    [ https://issues.apache.org/jira/browse/HBASE-17777?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16025909#comment-16025909

ramkrishna.s.vasudevan commented on HBASE-17777:

Seems the failures are not related to the patch. If some one could review it I could commit
this change.
Ping for reviews !! 
[~anoop.hbase], [~ted_yu].

> TestMemstoreLAB#testLABThreading runs too long for a small test
> ---------------------------------------------------------------
>                 Key: HBASE-17777
>                 URL: https://issues.apache.org/jira/browse/HBASE-17777
>             Project: HBase
>          Issue Type: Improvement
>          Components: test
>    Affects Versions: 2.0.0
>            Reporter: ramkrishna.s.vasudevan
>            Assignee: ramkrishna.s.vasudevan
>            Priority: Minor
>         Attachments: HBASE-17777_1.patch
> While working on ChunkCreator/ChunkMap found that the test in TestMSLAB#testLABThreading()
runs for almost 5 mins and the whole test is under smallTest category.
> The reason is that we are creating 35*2MB chunks from MSLAB. We try writing data to these
chunks until they are 50MB in size.
> And while verifying in order to check if the chunks are not overwritten/overlapped we
verify the content of the buffers.
> So we actually keep comparing 50MB buffer n  number of times. I suggest we change this
in a way that at max we create chunks whose size is totally at 1MB or may be even lesser and
write cells which are smaller in size. By doing this we can drastically reduce the run time
of this test. May be something less than 1 min.

This message was sent by Atlassian JIRA

View raw message