[jira] [Comment Edited] (IGNITE-11419) Memory leak after multiple restarts of server node within the same thread

["Pavel Vinokurov (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/IGNITE-11419?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16784180#comment-16784180 ] 

Pavel Vinokurov edited comment on IGNITE-11419 at 3/5/19 8:28 AM:
------------------------------------------------------------------

ConnectionManager declares the thread local variable initialized by the instance of anonymous class.Thus the thread local variable linked with the ConnectionManager. It leads to memory leak.
{code:java}
    /** Connection cache. */
    private final ThreadLocal<ThreadLocalObjectPool<H2ConnectionWrapper>.Reusable> threadConn =
        new ThreadLocal<ThreadLocalObjectPool<H2ConnectionWrapper>.Reusable>() {
        @Override public ThreadLocalObjectPool<H2ConnectionWrapper>.Reusable get() {
            ThreadLocalObjectPool<H2ConnectionWrapper>.Reusable reusable = super.get();
{code}



was (Author: pvinokurov):
ConnectionManager declares the thread local variable initialized by the instance of anonymous class.Thus the thread local variably linked with the ConnectionManager. It leads to memory leak.
{code:java}
    /** Connection cache. */
    private final ThreadLocal<ThreadLocalObjectPool<H2ConnectionWrapper>.Reusable> threadConn =
        new ThreadLocal<ThreadLocalObjectPool<H2ConnectionWrapper>.Reusable>() {
        @Override public ThreadLocalObjectPool<H2ConnectionWrapper>.Reusable get() {
            ThreadLocalObjectPool<H2ConnectionWrapper>.Reusable reusable = super.get();
{code}


> Memory leak after multiple restarts of server node within the same thread
> -------------------------------------------------------------------------
>
>                 Key: IGNITE-11419
>                 URL: https://issues.apache.org/jira/browse/IGNITE-11419
>             Project: Ignite
>          Issue Type: Bug
>          Components: sql
>    Affects Versions: 2.7
>            Reporter: Pavel Vinokurov
>            Priority: Minor
>          Time Spent: 10m
>  Remaining Estimate: 0h
>
> Multiple restarts of a server node with enabled persistence and 20 caches lead to OutOfMemory.



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)