[DIRSERVER-2176] Thread leak from LdapServer

• Type: Bug
• Status: Resolved
• Priority: Minor
• Resolution: Fixed
• Affects Version/s: 2.0.0-M21
• Fix Version/s: 2.0.0.AM27
• Component/s: core
• Assignee: Unassigned
• Reporter: Hrishikesh Gadre
• Created: 20/Dec/16 21:52
• Updated: 14/Aug/22 13:02
• Resolved: 14/Aug/22 13:02

Description

While using ApacheDS library to verify LDAP integration for Solr, I noticed a thread-leakage in the LdapServer code. Specifically the UnorderedThreadPoolExecutor added as part of LdapServer#start() method is not cleaned properly during the stop() method.

https://github.com/apache/directory-server/blob/7391447b3cee132bbcc9a023f435919025c28511/protocol-ldap/src/main/java/org/apache/directory/server/ldap/LdapServer.java#L525

The carrotsearch randomized testing library in Solr reports following stack-trace as part of thread leak detection,

[junit4] 2> Dec 20, 2016 1:32:31 PM com.carrotsearch.randomizedtesting.ThreadLeakControl checkThreadLeaks
[junit4] 2> SEVERE: 1 thread leaked from SUITE scope at org.apache.solr.security.hadoop.TestSolrCloudWithHadoopAuthPlugin:
[junit4] 2> 1) Thread[id=75, name=pool-4-thread-1, state=TIMED_WAITING, group=TGRP-TestSolrCloudWithHadoopAuthPlugin]
[junit4] 2> at sun.misc.Unsafe.park(Native Method)
[junit4] 2> at java.util.concurrent.locks.LockSupport.parkNanos(LockSupport.java:215)
[junit4] 2> at java.util.concurrent.locks.AbstractQueuedSynchronizer$ConditionObject.awaitNanos(AbstractQueuedSynchronizer.java:2078)
[junit4] 2> at java.util.concurrent.LinkedBlockingQueue.poll(LinkedBlockingQueue.java:467)
[junit4] 2> at org.apache.mina.filter.executor.UnorderedThreadPoolExecutor$Worker.fetchTask(UnorderedThreadPoolExecutor.java:462)
[junit4] 2> at org.apache.mina.filter.executor.UnorderedThreadPoolExecutor$Worker.run(UnorderedThreadPoolExecutor.java:414)
[junit4] 2> at java.lang.Thread.run(Thread.java:745)