[
https://issues.apache.org/jira/browse/FLINK-8893?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16391248#comment-16391248
]
Nico Kruber commented on FLINK-8893:
------------------------------------
I looked at the merged commits in Netty but the fix commit at
https://github.com/netty/netty/commit/c0fd74858a16cd3783e1bff0bb0e3da060c63971
claims this was caused by
https://github.com/netty/netty/commit/f18990a8a507d52fc40416d169db340105b10ec0
(reducing synchronization) which was initially merged into 4.0.29 and fixed by
4.0.33.
Since we are currently using 4.0.27, either the bug has been in Netty even
before the claimed commit, or the cause is somewhere else.
> NPE when netty try to allocate directBuffer
> -------------------------------------------
>
> Key: FLINK-8893
> URL: https://issues.apache.org/jira/browse/FLINK-8893
> Project: Flink
> Issue Type: Bug
> Components: Network
> Affects Versions: 1.3.2
> Reporter: aitozi
> Priority: Blocker
>
> Job failed with this exception
> {code:java}
> Caused by: java.lang.NullPointerException
> at io.netty.buffer.PoolChunk.initBufWithSubpage(PoolChunk.java:381)
> at io.netty.buffer.PoolChunk.initBufWithSubpage(PoolChunk.java:369)
> at io.netty.buffer.PoolArena.allocate(PoolArena.java:194)
> at io.netty.buffer.PoolArena.allocate(PoolArena.java:132)
> at
> io.netty.buffer.PooledByteBufAllocator.newDirectBuffer(PooledByteBufAllocator.java:271)
> at
> io.netty.buffer.AbstractByteBufAllocator.directBuffer(AbstractByteBufAllocator.java:155)
> at
> io.netty.buffer.AbstractByteBufAllocator.directBuffer(AbstractByteBufAllocator.java:146)
> at
> org.apache.flink.runtime.io.network.netty.NettyBufferPool.directBuffer(NettyBufferPool.java:278)
> at
> org.apache.flink.runtime.io.network.netty.NettyMessage.allocateBuffer(NettyMessage.java:72)
> at
> org.apache.flink.runtime.io.network.netty.NettyMessage.access$000(NettyMessage.java:50)
> at
> org.apache.flink.runtime.io.network.netty.NettyMessage$BufferResponse.write(NettyMessage.java:227)
> ... 26 more
> {code}
> After research, this may be caused by allocate from muti-thread.
> https://github.com/netty/netty/issues/4198
> https://github.com/netty/netty/pull/4388
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
