[jira] [Resolved] (HBASE-10668) TestExportSnapshot runs too long

["Andrew Purtell (JIRA)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/HBASE-10668?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Andrew Purtell resolved HBASE-10668.
------------------------------------

    Resolution: Done
      Assignee:     (was: Rekha Joshi)

Agreed. 

Also, I did a naive split of the test cases into two units, and the total running time exceeded by 100 seconds the time all cases require in a single unit. I could see by random sampling of call stacks with jstack that the difference was YARN minicluster related teardown. During test execution, if watching java processes with jps you might see the application master hanging around with no executing child for a while. We might be able to tune the YARN minicluster to get additional speedup if this becomes a problem.

> TestExportSnapshot runs too long
> --------------------------------
>
>                 Key: HBASE-10668
>                 URL: https://issues.apache.org/jira/browse/HBASE-10668
>             Project: HBase
>          Issue Type: Bug
>    Affects Versions: 0.98.0
>            Reporter: Andrew Purtell
>             Fix For: 0.98.1, 0.99.0
>
>
> 332 seconds
> {noformat}
> Forking command line: /bin/sh -c cd /data/src/hbase/hbase-server && /usr/lib/jvm/java-1.7.0.45-oracle-amd64/jre/bin/java -enableassertions -Xmx1900m -XX:MaxPermSize=100m -Djava.security.egd=file:/dev/./urandom -Djava.net.preferIPv4Stack=true -Djava.awt.headless=true -jar /data/src/hbase/hbase-server/target/surefire/surefirebooter1668068702110669265.jar /data/src/hbase/hbase-server/target/surefire/surefire5744357307851892501tmp /data/src/hbase/hbase-server/target/surefire/surefire_3661340119563945183029tmp
> Running org.apache.hadoop.hbase.snapshot.TestExportSnapshot
> Tests run: 7, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 332.421 sec
> {noformat}
> Slim down or split up.



--
This message was sent by Atlassian JIRA
(v6.2#6252)