[cisco-voip] Simulate calls

[Kevin Thorngren]

Hi Lasse,

I haven't used iperf and I am not sure how you used it to cause the CCM  
to have communications issues.  Short of purchasing something like the  
IXIA Voice test tool  - I am not familiar with anything that is  
freeware/shareware that provides SCCP calls simulation.

There are some things you can do to see if you are having issues with  
the ICCS communications between the CCM nodes.  First you can look in  
the Application Event log for SDLLinkOOS errors.  You can find some  
information on this message here:
http://www.cisco.com/en/US/products/sw/voicesw/ps556/ 
prod_error_message09186a0080307a94.html

and

http://www.cisco.com/en/US/products/sw/voicesw/ps556/ 
products_tech_note09186a0080111ac2.shtml


A packet capture of each CallManager server in question should quickly  
show you if there are network issues.  If you are seeing TCP  
Retransmissions between the two servers then you will need to track  
down the cause of the retransmissions.  Decoding these packets  
shouldn't be necessary.  Just need to make sure the TCP communications  
is working properly.  You can search CCO for "callmanager tcp port  
usage" and find there are docs for most versions.  In the 4.1 doc there  
is a section that outlines the various TCP/UDP ports used for all ICCS  
communications between the CCMs.

The CCM SRND has some information regarding ICCS and WAN connectivity.   
For example, this section discusses 4.x design info:
http://www.cisco.com/en/US/products/sw/voicesw/ps556/ 
products_implementation_design_guide_chapter09186a0080447510.html

The CCM/SDL contain some information but you would be better off using  
the packet captures.

Hope this helps to get you started.

Kevin

On Jul 6, 2006, at 11:21 AM, Lasse Kim Christiansen wrote:

> Hi all,
>
> I have some issues with a distributed callmanager cluster where i
> though i might have intracluster communications issues (or networking
> issues).
>
> To test it i ran iperf (to test load and jitter) between the two
> relevant callmanagers. When doing that (200 sessions with 12k each)
> both callmanager stopped responding to normal phone requests and all
> phones dropped off until we stopped iperf.
>
> The setup is running on a 100meg full duplex.
>
> My question is. Do anyone know how to do stress testing of a
> callmanager system like
> - Generate lots of call setups (signalling)
>
> And then verify when a callmanager cannot handle the load anymore ?
> At the same time do anyone know how to verify (with protocol analyzer
> or with trace files) if the intra-cluster communications actually
> work as expected ?
>
> Thanks in advance,
>
> Lasse,
>
> _______________________________________________
> cisco-voip mailing list
> cisco-voip at puck.nether.net
> https://puck.nether.net/mailman/listinfo/cisco-voip
>