List Info

Thread: Memory leak
Memory leak
country flaguser name
United States		 2007-07-23 14:33:47 

  


  

    


Hello,




We are using the latest jar for red5, and we are experiencing out of memory errors after about 4 hours of heavy use (with 2 gig heap).  We did a quick profile with jmap and saw large ConcurrentHashmaps, has anyone else seen this behavior?



Thanks,




Adrian






  

  
  
   
     
    

  

    Re:  Memory leak
  


  

     country flaguser name
United States		
     2007-07-24 15:30:38
  


  

    

  


  

    


Hi,





I believe we've found a bug.  




In the Red5 class(line 51):




private static ThreadLocal<WeakReference<IConnection>> connThreadLocal = new ThreadLocal<WeakReference<IConnection>>();




The theory is, under load, the threads get waited and occasionally this Iconnection gets garbage collected out from underneath due to the WeakReference.



When this object is null, ; a NPE exception is thrown and the connection is never removed from this Map in Scope.java(line 250)



private volatile Map<IClient, Set<IConnection>;> clients = new ConcurrentHashMap<IClient, Set<IConnection>;>();






Which leads to a memory leak. ; We wrapped this call in a try/catch and the problem has gone away.





Best,





Adrian








_____________________________________________ 


From:   Roston, Adrian - iFilm ; 


Sent: ;  Monday, July 23, 2007 12:34 PM


To: &nbsp;   'red5osflash.org'


Subject:&nbsp; &nbsp;   ;  Memory leak




Hello,




We are using the latest jar for red5, and we are experiencing out of memory errors after about 4 hours of heavy use (with 2 gig heap).&nbsp; We did a quick profile with jmap and saw large ConcurrentHashmaps, has anyone else seen this behavior?



Thanks,




Adrian






  

  
  
   
     
    

  

    Re:  Memory leak
  


  

     user name

     2007-07-24 15:49:39
  


  

    
Hi Adrian,

Roston, Adrian - iFilm schrieb:
> I believe we've found a bug. 
[...]
> Which leads to a memory leak.  We wrapped this call in
a try/catch and
> the problem has gone away.

thanks for pointing this out. Could you please describe a
bit more where
the NPE is thrown and in which cases?
Anyway, please create a patch about your try/catch change
against the
latest SVN trunk and add a ticket to jira.red5.org so we can
fix this
issue.

Joachim


_______________________________________________
Red5 mailing list
Red5osflash.org

http://osflash.org/mailman/listinfo/red5_osflash.org



  


  

    
  

  
  
   
     
    

  

    Re:  Memory leak
  


  

     country flaguser name
United States		
     2007-07-24 16:34:45
  


  

    
Sure...

Here's the stack trace


java.lang.NullPointerException
	at
org.red5.server.CoreHandler.connect(CoreHandler.java:77)
	at org.red5.server.Scope.connect(Scope.java:391)
	at org.red5.server.Scope.connect(Scope.java:388)
	at
org.red5.server.BaseConnection.connect(BaseConnection.java:2
64)
	at
org.red5.server.net.rtmp.RTMPConnection.connect(RTMPConnecti
on.java:241)
	at
org.red5.server.net.rtmp.RTMPMinaConnection.connect(RTMPMina
Connection.j
ava:67)
	at
org.red5.server.net.rtmp.RTMPHandler.onInvoke(RTMPHandler.ja
va:266)
	at
org.red5.server.net.rtmp.BaseRTMPHandler.messageReceived(Bas
eRTMPHandler
.java:143)
	at
org.red5.server.net.rtmp.RTMPMinaIoHandler.messageReceived(R
TMPMinaIoHan
dler.java:120)
	at
org.apache.mina.common.support.AbstractIoFilterChain$TailFil
ter.messageR
eceived(AbstractIoFilterChain.java:703)
	at
org.apache.mina.common.support.AbstractIoFilterChain.callNex
tMessageRece
ived(AbstractIoFilterChain.java:362)
	at
org.apache.mina.common.support.AbstractIoFilterChain.access$
1100(Abstrac
tIoFilterChain.java:54)
	at
org.apache.mina.common.support.AbstractIoFilterChain$EntryIm
pl$1.message
Received(AbstractIoFilterChain.java:800)
	at
org.apache.mina.filter.executor.ExecutorFilter.processEvent(
ExecutorFilt
er.java:247)
	at
org.apache.mina.filter.executor.ExecutorFilter$ProcessEvents
Runnable.run
(ExecutorFilter.java:307)
	at
java.util.concurrent.ThreadPoolExecutor$Worker.runTask(Threa
dPoolExecuto
r.java:885)
	at
java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoo
lExecutor.ja
va:907)
	at java.lang.Thread.run(Thread.java:619) 

The caveat to the fix we did is, this only prevents the
memory leak, but
does not solve the root of the problem.  Perhaps using a
harvesting
thread to do the removals will solve it, rather than using
WeakReferences and counting on the vm to do the cleanup?  I
opened the
ticket in JIRA.

--Adrian



-----Original Message-----
From: red5-bouncesosflash.org [mailto:red5-bouncesosflash.org] On
Behalf Of Joachim Bauch
Sent: Tuesday, July 24, 2007 1:50 PM
To: red5osflash.org
Subject: Re: [Red5] Memory leak

Hi Adrian,

Roston, Adrian - iFilm schrieb:
> I believe we've found a bug. 
[...]
> Which leads to a memory leak.  We wrapped this call in
a try/catch and

> the problem has gone away.

thanks for pointing this out. Could you please describe a
bit more where
the NPE is thrown and in which cases?
Anyway, please create a patch about your try/catch change
against the
latest SVN trunk and add a ticket to jira.red5.org so we can
fix this
issue.

Joachim


_______________________________________________
Red5 mailing list
Red5osflash.org

http://osflash.org/mailman/listinfo/red5_osflash.org

_______________________________________________
Red5 mailing list
Red5osflash.org

http://osflash.org/mailman/listinfo/red5_osflash.org



  


  

    
  

  
  
   
     
    






 [1-4]











   
 





about | contact  Other archives ( Real Estate discussion Medical topics )