thrift-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Mario Emmenlauer (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (THRIFT-4598) TBinaryProtocol.readMessageBegin() hangs forever in Java
Date Wed, 04 Jul 2018 15:26:00 GMT

    [ https://issues.apache.org/jira/browse/THRIFT-4598?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16532879#comment-16532879
] 

Mario Emmenlauer commented on THRIFT-4598:
------------------------------------------

Thanks for this very helpful insight! I was indeed using {code}TServer::stop(){code} instead
of {code}TThreadPoolServer::stop(){code}
This seems to have a significant difference. The server exits cleanly now after using the
latter. For me this fixes the issue, but maybe we can leave the report open until TServer
is adjusted accordingly?

> TBinaryProtocol.readMessageBegin() hangs forever in Java
> --------------------------------------------------------
>
>                 Key: THRIFT-4598
>                 URL: https://issues.apache.org/jira/browse/THRIFT-4598
>             Project: Thrift
>          Issue Type: Bug
>          Components: Java - Library
>    Affects Versions: 0.11.0
>         Environment: Ubuntu Linux 16.04 x86_64
>            Reporter: Mario Emmenlauer
>            Priority: Major
>
> I run some internal tests for the robustness of our thrift Java server. One of the robustness
tests is to try to connect to the secure server socket with an insecure client and vice versa.
This may seem a slightly diabolic test but it can be a simple user error to forget enabling
or disabling encryption. So I think thrift should handle this gracefully.
> However it seems that this test leads to an infinitely hanging server thread! The thread
blocks the jvm (Java 1.8.0_161 x86_64) from ending. I can see in jstack that the main method
ends and there are almost no threads left, except some garbage collectors and finalizers
plus org.apache.thrift.server.TThreadPoolServer.
> Here is the call stack of the thrift server thread:
> {code:java}
> java.lang.Thread.State: RUNNABLE
> at java.net.SocketInputStream.socketRead0(Native Method)
> at java.net.SocketInputStream.socketRead(SocketInputStream.java:116)
> at java.net.SocketInputStream.read(SocketInputStream.java:171)
> at java.net.SocketInputStream.read(SocketInputStream.java:141)
> at java.io.BufferedInputStream.fill(BufferedInputStream.java:246)
> at java.io.BufferedInputStream.read1(BufferedInputStream.java:286)
> at java.io.BufferedInputStream.read(BufferedInputStream.java:345)
> - locked <0x000000067b390fd8> (a java.io.BufferedInputStream)
> at org.apache.thrift.transport.TIOStreamTransport.read(TIOStreamTransport.java:127)
> at org.apache.thrift.transport.TTransport.readAll(TTransport.java:86)
> at org.apache.thrift.protocol.TBinaryProtocol.readAll(TBinaryProtocol.java:425)
> at org.apache.thrift.protocol.TBinaryProtocol.readI32(TBinaryProtocol.java:321)
> at org.apache.thrift.protocol.TBinaryProtocol.readMessageBegin(TBinaryProtocol.java:225)
> at org.apache.thrift.TBaseProcessor.process(TBaseProcessor.java:27)
> at org.apache.thrift.server.TThreadPoolServer$WorkerProcess.run(TThreadPoolServer.java:310)
> at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
> at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
> at java.lang.Thread.run(Thread.java:748)
> {code}
> I already reduced the server socket timeout to 30sec and the requestTimeout to 10sec.
Is there anything else I can do to timeout readMessageBegin()?



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

Mime
View raw message