[ 
https://issues.apache.org/jira/browse/GEODE-3314?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16118721#comment-16118721
 ] 

ASF GitHub Bot commented on GEODE-3314:
---------------------------------------

Github user pivotal-amurmann commented on a diff in the pull request:

    https://github.com/apache/geode/pull/683#discussion_r131977004
  
    --- Diff: 
geode-core/src/main/java/org/apache/geode/distributed/internal/locks/DLockService.java
 ---
    @@ -1433,29 +1412,7 @@ public boolean lockInterruptibly(final Object name, 
final long waitTimeMillis,
             int lockId = -1;
             incActiveLocks();
     
    -        int loopCount = 0;
             while (keepTrying) {
    --- End diff --
    
    This `keepTrying` thing is very unclear because it hides what the actual 
decision is based on. It seems like we keep trying until either the `waitLimit` 
has been exceeded or the lock has been acquired. Could we split `keepTrying` 
out into `boolean lockAcquired` and `boolean waitLimitExceeded`? Maybe the 
check for `waitTimeLimit` doesn't even need a variable but could be checked by 
the `while`?


> OldGen heap usage grows without increase in data size, large number of 
> DLockTokens
> ----------------------------------------------------------------------------------
>
>                 Key: GEODE-3314
>                 URL: https://issues.apache.org/jira/browse/GEODE-3314
>             Project: Geode
>          Issue Type: Bug
>          Components: distributed lock service
>            Reporter: Udo Kohlmeyer
>            Assignee: Galen O'Sullivan
>             Fix For: 1.2.1
>
>
> When using Global Replicate regions, the DLockTokens are not correctly 
> cleaned up, causing a memory leak.



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)

Reply via email to