Mailing-List: contact issues-help@hbase.apache.org; run by ezmlm
Precedence: bulk
Date: Wed, 12 Dec 2012 04:15:22 +0000 (UTC)
From: "Ted Yu (JIRA)" <jira@apache.org>
To: issues@hbase.apache.org
Message-ID: <JIRA.12617239.1353541282073.21395.1355285722633@arcas>
In-Reply-To: <JIRA.12617239.1353541282073@arcas>
References: <JIRA.12617239.1353541282073@arcas>
Subject: [jira] [Commented] (HBASE-7205) Coprocessor classloader is
 replicated for all regions in the HRegionServer
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit


    [ https://issues.apache.org/jira/browse/HBASE-7205?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13529614#comment-13529614 ] 

Ted Yu commented on HBASE-7205:
-------------------------------

Did some experiment. First, as http://docs.oracle.com/javase/1.5.0/docs/api/java/lang/management/ManagementFactory.html#getGarbageCollectorMXBeans%28%29 says, we may need to deal with several GarbageCollectorMXBean's.

Here is what I came up with:
{code}
    admin.disableTable(htd.getName());
    admin.deleteTable(htd.getName());
    List<GarbageCollectorMXBean> mxbeans = ManagementFactory.getGarbageCollectorMXBeans();
    LOG.info("There are " + mxbeans.size() + " MXBeans");
    GarbageCollectorMXBean mxbean = mxbeans.get(0);
    long initialCollectionCount = mxbean.getCollectionCount();
    long startTime = System.currentTimeMillis();
    System.gc();
    while (mxbean.getCollectionCount() == initialCollectionCount &&
        System.currentTimeMillis() - startTime < 30000) {
      Thread.sleep(10);
    }
    for (Path p : CoprocessorHost.classLoadersCache.keySet()) {
      ClassLoader cl = CoprocessorHost.classLoadersCache.get(p);
      assertTrue(" " + cl, cl == null);
    }
{code}
After waiting for 30 seconds, the assertion fails:
{code}
testClassLoadingFromHDFS(org.apache.hadoop.hbase.coprocessor.TestClassLoading)  Time elapsed: 39.79 sec  <<< FAILURE!
java.lang.AssertionError:  org.apache.hadoop.hbase.coprocessor.CoprocessorClassLoader@24e7cb66
  at org.junit.Assert.fail(Assert.java:93)
  at org.junit.Assert.assertTrue(Assert.java:43)
  at org.apache.hadoop.hbase.coprocessor.TestClassLoading.testClassLoadingFromHDFS(TestClassLoading.java:339)
{code}
I don't know how long the test should wait for GC to kick in.
                
> Coprocessor classloader is replicated for all regions in the HRegionServer
> --------------------------------------------------------------------------
>
>                 Key: HBASE-7205
>                 URL: https://issues.apache.org/jira/browse/HBASE-7205
>             Project: HBase
>          Issue Type: Bug
>          Components: Coprocessors
>    Affects Versions: 0.92.2, 0.94.2
>            Reporter: Adrian Muraru
>            Assignee: Ted Yu
>            Priority: Critical
>             Fix For: 0.96.0, 0.94.4
>
>         Attachments: 7205-0.94.txt, 7205-v10.txt, 7205-v1.txt, 7205-v3.txt, 7205-v4.txt, 7205-v5.txt, 7205-v6.txt, 7205-v7.txt, 7205-v8.txt, 7205-v9.txt, HBASE-7205_v2.patch
>
>
> HBASE-6308 introduced a new custom CoprocessorClassLoader to load the coprocessor classes and a new instance of this CL is created for each single HRegion opened. This leads to OOME-PermGen when the number of regions go above hundres / region server. 
> Having the table coprocessor jailed in a separate classloader is good however we should create only one for all regions of a table in each HRS.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira