Hello Mike

Thanks for your answer. I tried with '-t u1' and I get the same error.
I also tried to run -mp 1024 and I added net.ipv4.ip_local_port_range
= 1024 65000 to /etc/sysctl.conf, and then sysctl -p /etc/sysctl.conf.
In my xml I use [auto_media_port], which now starts from 1024 and I
should have enough ports.

This still did not fix the problem.



2009/4/15 Mike Ayers <mike_ay...@tvworks.com>:
>> From: catalina oancea [mailto:catalina.oan...@gmail.com]
>> Sent: Wednesday, April 15, 2009 7:33 AM
>
>> The problem also occurs if running as uas.
>>
>> Has anybody tried sipp with rtp for all calls with 200 calls?
>
>        Have you checked `netstat -na` (or equivalent) to see if you are 
> running out of ports?  200 ports per second times 2 minutes (typical) 
> TIME_WAIT yields 24,000 ports in churn, which tends to increase as resource 
> usage increases and the OS is slower in getting around to cleaning them up.
>
>        To test this theory, change '-t un' to '-t u1', just for test purposes.
>
>        A possible workaround is to add multiple IP aliases to the addresses 
> (make sure they're not in use!) and use the injection file to bind the call 
> source to different aliases, thus slowing down the churn rate on each address.
>
>
>        HTH,
>
> Mike
>

------------------------------------------------------------------------------
Stay on top of everything new and different, both inside and 
around Java (TM) technology - register by April 22, and save
$200 on the JavaOne (SM) conference, June 2-5, 2009, San Francisco.
300 plus technical and hands-on sessions. Register today. 
Use priority code J9JMT32. http://p.sf.net/sfu/p
_______________________________________________
Sipp-users mailing list
Sipp-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sipp-users

Reply via email to