flink-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "ASF GitHub Bot (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (FLINK-9860) Netty resource leak on receiver side
Date Wed, 18 Jul 2018 15:30:00 GMT

    [ https://issues.apache.org/jira/browse/FLINK-9860?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16547968#comment-16547968
] 

ASF GitHub Bot commented on FLINK-9860:
---------------------------------------

Github user zentol commented on a diff in the pull request:

    https://github.com/apache/flink/pull/6363#discussion_r203422176
  
    --- Diff: flink-runtime/src/test/java/org/apache/flink/runtime/io/network/netty/NettyLeakDetectionResource.java
---
    @@ -0,0 +1,84 @@
    +/*
    + * Licensed to the Apache Software Foundation (ASF) under one
    + * or more contributor license agreements.  See the NOTICE file
    + * distributed with this work for additional information
    + * regarding copyright ownership.  The ASF licenses this file
    + * to you under the Apache License, Version 2.0 (the
    + * "License"); you may not use this file except in compliance
    + * with the License.  You may obtain a copy of the License at
    + *
    + *     http://www.apache.org/licenses/LICENSE-2.0
    + *
    + * Unless required by applicable law or agreed to in writing, software
    + * distributed under the License is distributed on an "AS IS" BASIS,
    + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
    + * See the License for the specific language governing permissions and
    + * limitations under the License.
    + */
    +
    +package org.apache.flink.runtime.io.network.netty;
    +
    +import org.apache.flink.shaded.netty4.io.netty.util.ResourceLeakDetector;
    +import org.apache.flink.shaded.netty4.io.netty.util.ResourceLeakDetectorFactory;
    +
    +import org.junit.rules.ExternalResource;
    +
    +import static org.junit.Assert.fail;
    +
    +/**
    + * JUnit resource to fail with an assertion when Netty detects a resource leak (only
with
    + * <tt>ERROR</tt> logging enabled for
    + * <tt>org.apache.flink.shaded.netty4.io.netty.util.ResourceLeakDetector</tt>).
    + *
    + * <p>This should be used in a class rule:
    + * <pre>{@code
    + * @literal @ClassRule
    + *  public static final NettyLeakDetectionResource LEAK_DETECTION = new NettyLeakDetectionResource();
    + * }</pre>
    + */
    +public class NettyLeakDetectionResource extends ExternalResource {
    +	private static ResourceLeakDetectorFactory previousLeakDetector;
    +	private static ResourceLeakDetector.Level previousLeakDetectorLevel;
    +
    +	@Override
    +	protected void before() {
    +		previousLeakDetector = ResourceLeakDetectorFactory.instance();
    +		previousLeakDetectorLevel = ResourceLeakDetector.getLevel();
    +
    +		ResourceLeakDetector.setLevel(ResourceLeakDetector.Level.PARANOID);
    +		ResourceLeakDetectorFactory.setResourceLeakDetectorFactory(new FailingResourceLeakDetectorFactory());
    --- End diff --
    
    so this isn't something we necessarily have to deal with now, but if multiple tests that
use this resource run in parallel we might not reset to the correct factory/level.
    
    Effectively what we need is some kinda ref-counting so that only the first resource modifies
the level and factory, and only the last resource reset them. :/
    



> Netty resource leak on receiver side
> ------------------------------------
>
>                 Key: FLINK-9860
>                 URL: https://issues.apache.org/jira/browse/FLINK-9860
>             Project: Flink
>          Issue Type: Bug
>          Components: Network
>    Affects Versions: 1.5.1, 1.6.0
>            Reporter: Till Rohrmann
>            Assignee: Nico Kruber
>            Priority: Blocker
>              Labels: pull-request-available, test-stability
>             Fix For: 1.5.2, 1.6.0
>
>
> The Hadoop-free Wordcount end-to-end test fails with the following exception:
> {code}
> ERROR org.apache.flink.shaded.netty4.io.netty.util.ResourceLeakDetector  - LEAK: ByteBuf.release()
was not called before it's garbage-collected. See http://netty.io/wiki/reference-counted-objects.html
for more information.
> Recent access records: 
> Created at:
> 	org.apache.flink.shaded.netty4.io.netty.buffer.PooledByteBufAllocator.newDirectBuffer(PooledByteBufAllocator.java:331)
> 	org.apache.flink.shaded.netty4.io.netty.buffer.AbstractByteBufAllocator.directBuffer(AbstractByteBufAllocator.java:185)
> 	org.apache.flink.shaded.netty4.io.netty.buffer.AbstractByteBufAllocator.directBuffer(AbstractByteBufAllocator.java:176)
> 	org.apache.flink.shaded.netty4.io.netty.buffer.AbstractByteBufAllocator.ioBuffer(AbstractByteBufAllocator.java:137)
> 	org.apache.flink.shaded.netty4.io.netty.channel.DefaultMaxMessagesRecvByteBufAllocator$MaxMessageHandle.allocate(DefaultMaxMessagesRecvByteBufAllocator.java:114)
> 	org.apache.flink.shaded.netty4.io.netty.channel.nio.AbstractNioByteChannel$NioByteUnsafe.read(AbstractNioByteChannel.java:147)
> 	org.apache.flink.shaded.netty4.io.netty.channel.nio.NioEventLoop.processSelectedKey(NioEventLoop.java:645)
> 	org.apache.flink.shaded.netty4.io.netty.channel.nio.NioEventLoop.processSelectedKeysOptimized(NioEventLoop.java:580)
> 	org.apache.flink.shaded.netty4.io.netty.channel.nio.NioEventLoop.processSelectedKeys(NioEventLoop.java:497)
> 	org.apache.flink.shaded.netty4.io.netty.channel.nio.NioEventLoop.run(NioEventLoop.java:459)
> 	org.apache.flink.shaded.netty4.io.netty.util.concurrent.SingleThreadEventExecutor$5.run(SingleThreadEventExecutor.java:884)
> 	org.apache.flink.shaded.netty4.io.netty.util.concurrent.FastThreadLocalRunnable.run(FastThreadLocalRunnable.java:30)
> {code}
> We might have a resource leak on the receiving side of our network stack.
> https://api.travis-ci.org/v3/job/404225956/log.txt



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

Mime
View raw message