File handles leak

[GoogleCodeExporter]

What steps will reproduce the problem?
1. To observe the problem monitor the number of open files by java process:
lsof -p PROCPID | wc -l, where PROCPID is the pid of the java process
running red5. Do this before and after calling. If server is used
intensively it starts throwing exceptions with message "Too many open
files" and practically stops functioning.

What is the expected output? What do you see instead?
It is expected after each call the number of open files (file handles) to
be restored to the number before the call. The number of available file
handles is spent quickly and the red5/red5phone becomes unusable. The
process starts throwing exceptions with message "Too many open files".

What version of the product are you using? On what operating system?
Operating system is Debian / GNU Linux 2.6. Java 1.5 and Java 6 produce the
same result. Using Red5 0.8 and Red5 0.9RC1 with latest sip.zip builds.

Please provide any additional information below.
While the number of limited file handles (sockets, files, threads) can
increase the time before the system becomes unusable, it does not solve the
stability problem. Practically, on every call the java process opens around
15-20 file handles and restores only 2 after call hangup. If a server is
more intensively used the system may become unusable within minutes.
Setting socket timeout value to a lower number (defaults to 60) also
doesn't help (as suggested for projects that use Mina for NIO).

Original issue reported on code.google.com by stoya...@gmail.com on 16 Nov 2009 at 3:29

[GoogleCodeExporter]

This leak is now fixed with Red 5 ver 0.9.0. However using red5phone with 0.9.0 
has 
the issue #89.

Original comment by stoya...@gmail.com on 19 Feb 2010 at 4:42

[GoogleCodeExporter]

Actually it is very important issue. As stoyanov said, your red5phone would be
unusable after using it a while. And as far as red5phone is not compatible with 
"Red5
0.9.0 Final" or later this problem seems to not have a solution. Red5phone team
should release the fix.
If anyone fixes this bug just let us know.

Original comment by mollamus...@gmail.com on 26 Mar 2010 at 9:02

[GoogleCodeExporter]

Thanks for raising this, mollamusaoglu. I think many people are looking for a 
solution. I 
myself will look deeper inside to, once I find some time. I will reply if I 
find anything.
Let's hope it will be fixed soon.

Original comment by stoya...@gmail.com on 26 Mar 2010 at 10:35