Re: Re: 'Non closed socket' bug

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Hello

Very simple situation:
 Origination side = > GNUGK = > Termination side

Sample origination side is soft-switch that can do hunting
and termination side is overloaded. After Q931-setup
termination side bring up fast start logical channels
and immidiatelly reject call with ReleaseComplete with
NoChannel cause.

All 3 points connected over fast connection and have
very small delay. Sample origination softswitch side
will make 5 attempts with Q931 setup's before stop
hunting for single call.

In this case GNUGK will recieve 5 calls immidiatelly
one after other with same phone number / CRV / CallID /
call-references !!!!

It will generate sokets leak and memory leak because GNUGK
is not delete calls fast after call was end. As for me
it must do it immidiately. It is very possible that on
network with short delay other attempt will got faster
then GnuGK will done internal destruction process for
same attempt that it was got sample 10 ms ago.
As result you will got memory and socket leaks.
It is not depend on number of CallSignalHandler - this
bug work very well with single call signal handler.

As result it will not work stable and will grow in memory.

May be I am wrong somewhere...
But this is my own expirience from GnuGK and this problem.

Ivanov


k> Hi again, Michal.
k> What arbitrary delay you talk about?
k> Yesterday one call placed at 18:38 Msk and ended at 18:41 Msk was closed by 
k> gnugk only at 20:03 Msk.
k> And logical channel between gnugk and terminating gw was not closed.
k> Among other things this phenomenon results in occurrence superfluous IP 
k> traffic. And I could not understand, why January, 1-st at rather small amount 
k> billing minutes IP traffic of time in 4 has exceeded usual IP traffic (on a 
k> parity)!
k> Excuse me for whining. You are not obliged to pay attention to this problem. 
k> But to pretend, that it does not exist, simply strange!
k> Igor.

k> В письме от 2 Январь 2004 18:21 вы написали:
>> There is nothing wrong with the open sockets. These can be deleted
>> with some arbitrary delay and should not affect other calls.




-------------------------------------------------------
This SF.net email is sponsored by: IBM Linux Tutorials.
Become an expert in LINUX or just sharpen your skills.  Sign up for IBM's
Free Linux Tutorials.  Learn everything from the bash shell to sys admin.
Click now! http://ads.osdn.com/?ad_id78&alloc_id371&opлk
_______________________________________________
List: Openh323gk-users@lists.sourceforge.net
Archive: http://sourceforge.net/mailarchive/forum.php?forum_id┘49
Homepage: http://www.gnugk.org/


[Index of Archives]     [SIP]     [Open H.323]     [Gnu Gatekeeper]     [Asterisk PBX]     [ISDN Cause Codes]     [Yosemite News]

  Powered by Linux