[
https://issues.apache.org/jira/browse/HBASE-15697?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15257805#comment-15257805
]
ramkrishna.s.vasudevan commented on HBASE-15697:
------------------------------------------------
The first difference is that in trunk the testCount is 50 whereas in branch-1
it is 100. So after HBASE-14970 the number of files will be more and though
compaction happens there is nothing to clear them as TestHRegion does not allow
the compaction discharge handler to run.
If we change the testcount to 50 as in trunk then the test case runs fine. But
the better fix is that we need to ensure that compaction runs frequently and
they are getting cleared.
Testcase 'testWritesWhileGetting' is the culprit and there is already a clear
comment saying
{code}
// This test flushes constantly and can cause many files to be created,
// possibly
// extending over the ulimit. Make sure compactions are aggressive in
// reducing
// the number of HFiles created.
{code}
> Excessive TestHRegion running time on branch-1
> ----------------------------------------------
>
> Key: HBASE-15697
> URL: https://issues.apache.org/jira/browse/HBASE-15697
> Project: HBase
> Issue Type: Bug
> Affects Versions: 1.3.0
> Reporter: Andrew Purtell
> Assignee: ramkrishna.s.vasudevan
>
> On my dev box TestHRegion takes about 90 seconds to complete in master and
> about 60 seconds in 0.98, but about 370 seconds in branch-1. Furthermore
> TestHRegion in branch-1 blew past my open files ulimit. I had to raise it
> from default in order for the unit to complete at all.
> I am going to bisect the recent history of branch-1 in search of a culprit
> and report back.
> {panel:title=master}
> Running org.apache.hadoop.hbase.regionserver.TestHRegion
> Tests run: 102, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 87.299 sec
> - in org.apache.hadoop.hbase.regionserver.TestHRegion
> Running org.apache.hadoop.hbase.regionserver.TestHRegion
> Tests run: 102, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 91.529 sec
> - in org.apache.hadoop.hbase.regionserver.TestHRegion
> Running org.apache.hadoop.hbase.regionserver.TestHRegion
> Tests run: 102, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 89.23 sec -
> in org.apache.hadoop.hbase.regionserver.TestHRegion
> {panel}
> {panel:title=branch-1}
> Running org.apache.hadoop.hbase.regionserver.TestHRegion
> Tests run: 102, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 368.868 sec
> - in org.apache.hadoop.hbase.regionserver.TestHRegion
> Running org.apache.hadoop.hbase.regionserver.TestHRegion
> Tests run: 102, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 366.203 sec
> - in org.apache.hadoop.hbase.regionserver.TestHRegion
> Running org.apache.hadoop.hbase.regionserver.TestHRegion
> Tests run: 102, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 345.806 sec
> - in org.apache.hadoop.hbase.regionserver.TestHRegion
> {panel}
> {panel:title=0.98}
> Running org.apache.hadoop.hbase.regionserver.TestHRegion
> Tests run: 90, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 61.038 sec -
> in org.apache.hadoop.hbase.regionserver.TestHRegion
> Running org.apache.hadoop.hbase.regionserver.TestHRegion
> Tests run: 90, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 56.382 sec -
> in org.apache.hadoop.hbase.regionserver.TestHRegion
> Running org.apache.hadoop.hbase.regionserver.TestHRegion
> Tests run: 90, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 63.509 sec -
> in org.apache.hadoop.hbase.regionserver.TestHRegion
> {panel}
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
