[ 
https://issues.apache.org/jira/browse/THRIFT-4598?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16533534#comment-16533534
 ] 

James E. King III commented on THRIFT-4598:
-------------------------------------------

Ahh, I get it now - sorry I didn't see this before.  As part of the work on the 
C++ side I modified the library so that when you call stop() on the TServer, 
the implementation can interrupt ALL the client readers.  My guess is the Java 
side does not have that feature, which means the server is essentially held 
hostage by any connected client.  See THRIFT-3447.  This is a common problem, 
also see THRIFT-2696 and THRIFT-3590, THRIFT-4500.  But really, it sounds like 
this is now a duplicate of THRIFT-4252 which you previously reported. :)

> 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
>            Assignee: James E. King III
>            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)

Reply via email to