db-derby-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Kathey Marsden (JIRA)" <j...@apache.org>
Subject [jira] Updated: (DERBY-3869) intermittent hang pinging server on Linux
Date Tue, 23 Sep 2008 20:07:44 GMT

     [ https://issues.apache.org/jira/browse/DERBY-3869?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]

Kathey Marsden updated DERBY-3869:
----------------------------------

    Attachment: derby-3869_diff.txt

So I think I understand why we get the hang.  The ClientThread does not actually get interupted
when clientThread.interrupt() is called.  This is because accept() is not one of the calls
that gets interrupted with Thread.interrupt().  So the clientThreead sticks around and accepts
the ping request.  Even after the server socket is closed that socket remains open.  Attached
is my attempt to fix the problem, which closes the clientSocket if the server is shutting
down.  This rectifies the hang.  The ping request may get an insufficient reply data message
but I think that is appropriate if the server is shutting down.    It is certainly better
than a hang.


> intermittent hang pinging  server on Linux
> ------------------------------------------
>
>                 Key: DERBY-3869
>                 URL: https://issues.apache.org/jira/browse/DERBY-3869
>             Project: Derby
>          Issue Type: Bug
>          Components: Network Server
>    Affects Versions: 10.5.0.0
>         Environment: java version "1.6.0"
> Java(TM) SE Runtime Environment (build pxi3260sr1-20080416_01(SR1))
> IBM J9 VM (build 2.4, J2RE 1.6.0 IBM J9 2.4 Linux x86-32 jvmxi3260-20080415_1876
> J9VM - 20080415_018762_lHdSMr
> JIT  - r9_20080415_1520
> GC   - 20080415_AA)
> JCL  - 20080412_01
>            Reporter: Kathey Marsden
>            Assignee: Kathey Marsden
>         Attachments: derby-3869_diff.txt, derby-3869_sleep_to_force_hang_diff.txt, derby.log,
javacore.20080903.183815.7684.0001.txt, Repro.java, TryHang.java
>
>
>  am looking at a intermittent hang with IBM 1.6 on Linux with the ping 
> command.  I am not entirely sure it is a jvm issue, but I have not been 
> able to reproduce the hang with other jvms.  
> The trace is
> 3XMTHREADINFO      "main" TID:0x08072500, j9thread_t:0x08057AF4, state:R, prio=5
> 3XMTHREADINFO1            (native thread ID:0x1E05, native priority:0x5, native policy:UNKNOWN)
> 4XESTACKTRACE          at java/net/SocketInputStream.socketRead0(Native Method)
> 4XESTACKTRACE          at java/net/SocketInputStream.read(SocketInputStream.java:140)
> 4XESTACKTRACE          at java/net/SocketInputStream.read(SocketInputStream.java:101)
> 4XESTACKTRACE          at org/apache/derby/impl/drda/NetworkServerControlImpl.fillReplyBuffer(NetworkServerControlImpl.java:2764)
> 4XESTACKTRACE          at org/apache/derby/impl/drda/NetworkServerControlImpl.readResult(NetworkServerControlImpl.java:2708)
> 4XESTACKTRACE          at org/apache/derby/impl/drda/NetworkServerControlImpl.pingWithNoOpen(NetworkServerControlImpl.java:1169)
> 4XESTACKTRACE          at org/apache/derby/impl/drda/NetworkServerControlImpl.ping(NetworkServerControlImpl.java:1144(Compiled
Code))
> 4XESTACKTRACE          at org/apache/derby/drda/NetworkServerControl.ping(NetworkServerControl.java:395(Compiled
Code))
> 4XESTACKTRACE          at Repro.pingForServerUp(Repro.java:38(Compiled Code))
> 4XESTACKTRACE          at Repro.startAndShutdown(Repro.java:20)
> The client has sent the ping, but there is no corresponding session on 
> the server side to process the
> command. The full thread dump is in.
> javacore.20080903.183815.7684.0001.txt
> The  program Repro.java shows the problem. It repeatedly starts the server, pings until
it 
> comes up, and then shuts down.
> In the derby.log I see a startup error, that the address is already in 
> use, so presumably the shutdown is not complete before we start the 
> server and then perhaps it shuts down mid ping causing the hang?
> 2008-09-04 01:37:51.048 GMT : Could not listen on port 1527 on host 127.0.0.1:
>  java.net.BindException: Address already in use
> An exception was thrown during network server startup. DRDA_ListenPort.S:Could not listen
on port 1527 on host 127.0.0.1:
>  java.net.BindException: Address already in use
> java.lang.reflect.InvocationTargetException
> 		 at sun.reflect.GeneratedMethodAccessor3.invoke(Unknown Source)
> 		 at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:37)
> 		 at java.lang.reflect.Method.invoke(Method.java:599)
> 		 at org.apache.derby.iapi.jdbc.DRDAServerStarter.run(DRDAServerStarter.java:236)
> 		 at java.lang.Thread.run(Thread.java:735)
> Caused by: java.lang.Exception: DRDA_ListenPort.S:Could not listen on port 1527 on host
127.0.0.1:
>  java.net.BindException: Address already in use
> 		 at java.lang.Throwable.<init>(Throwable.java:67)
> 		 at org.apache.derby.impl.drda.NetworkServerControlImpl.consolePropertyMessageWork(NetworkServerControlImpl.java:3179)
> 		 at org.apache.derby.impl.drda.NetworkServerControlImpl.consolePropertyMessage(NetworkServerControlImpl.java:1861)
> 		 at org.apache.derby.impl.drda.NetworkServerControlImpl.blockingStart(NetworkServerControlImpl.java:731)
> 		 ... 5 more
> Full log is attached as derby.log

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


Mime
View raw message