I always face the same problem and I don't know why either....
>From what I noticed in my tests, the situation is the following: sometimes
when SIPp doesn't receive all the messages it is expecting to terminate the
scenario with no errors, it keeps these calls opened forever!! When this
situation occurs, SIPp respects the maximum number of sockets opened and
starts to reduce the call rate until it reaches 0 when it is running but
not sending any messages.... I think it is a bug since it should timeout
the expecting message, terminate the call and free the socket....
Does anyone knows the explanation for this?? Once I tried to investigate
the sources to understand what was happening, but couldn't find any
hint.....
On Sat, Apr 28, 2012 at 12:19 PM, giuseppe cofano <g.cofan...@gmail.com>wrote:
> Hello,
>
> I'm doing a stress test to a SIP proxy server to analyse the performance
> of several overload control algorithms I've implemented.
>
> The network's topology I'm using is the following: on the same host, a
> Centrino Duo with 2 GB Ram, I'm running uas and uac, while on the second
> one (Pentium Three with 512 MB Ram), linked through a Gb crossover cable,
> I'm running my modified proxy server. I have attached SIPp xml scenario
> files and bash scripts used to start tests, where you can see the options
> I've set to run SIPp instances. I'm using Openser version 1.4 for my proxy,
> modified with the introduction of a queue for coming Invite messages, that
> are managed according to the overload control mechanism.
>
> The problem I'm running into during my tests is that SIPp uac can't keep
> generating the target call-rate in situations of overload. I mean that it
> can easily generate high call-rate when the proxy doesn't suffer from
> overload problems, but when openser comes to the overload condition, for
> very low rates too (incoming rate at which overload verifies depends on the
> tested algorithm), the real call-rate becomes very unsteady and meanly
> significantly lower than the nominal call-rate value I've set. The cpu
> usage oh the host running SIPp doesn't exceed 50% and the set limit for
> simultaneously calls (200000) is much higher than necessary. Is there
> anything wrong in what I've done or what else?
>
> Thanks in advance.
>
> Regards,
> Giuseppe
>
>
> ------------------------------------------------------------------------------
> Live Security Virtual Conference
> Exclusive live event will cover all the ways today's security and
> threat landscape has changed and how IT managers can respond. Discussions
> will include endpoint security, mobile security and the latest in malware
> threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
> _______________________________________________
> Sipp-users mailing list
> Sipp-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/sipp-users
>
>
------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
Sipp-users mailing list
Sipp-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sipp-users
