08-21-2001 03:14 AM - edited 03-12-2019 12:19 PM
Hi,
I got a Problem:
Loosing often 1-3 seconds of speech at the beginning of voice calls made from PSTN to ip phones using an IOS gateways (3640 & E1 )
IP_phone-----cm-------IOS GW---pstn
<<<<<----------call------<<<<<<
Does anyone know a solution for this issue ?
Markus
08-21-2001 09:10 AM
I had the same problem and solved it by changing all call codecs between all Regions to g.729.
I have Centralized Call Processing model, and the remote GW was setting up the call as g.729. But when the IP phone picked up, the codec changed to g.711 because of the Region settings and that seemed to hose up the initial conversation. Maybe Dave can explain it better.
08-21-2001 10:44 AM
Can you post your GW config? I experienced this problem with an H323 GW config when I turned comfort noise OFF. I turned it back on to remedy the situation.
09-18-2001 12:32 PM
We also had this problem and we tried a number of things to clean it up.
First we went thru and made sure voip had priority
on every switch port and router port.
Second, there was one parameter that we changed in Call Manager. There is a parameter DTSilenceFlag that is set to false by default. According to the local Cisco techs, this has something to do with how a call gets setup. We changed it to True.
Last, we found a memory leak bug that was happening
on call mgr. Call manager was eating up any/all memory that it could find. We got that resolved and our delay went away.
Mike
09-18-2001 02:09 PM
Hi Mike,
How did you resolve the memory leak bug on Callmanager ? Can you share with us ? Thank you very much.
Duc Vu
09-21-2001 12:34 PM
First off, we rebooted call manager.
Second, there are a couple bugs that TAC is aware of that deal with the memory leaks. It is
related to a 6608 blade and having some of the T1 ports not being used and constantly attempting to register with the CM. Each failed registration attempt causes abit of memory to be lost.
09-26-2001 06:40 AM
Message for MIKE LIKER,
Mike,
I am having a hard time also keeping my end-users from lining up at my desk with b-ball bats... they are loosing patience with the system and the delay problems I am experiencing sounds alot like what you have gone thru.
Router/GW- Dial-peer ip precedence 5
C6509/c3524 - Where do check the priority on switches?
CM3.08 DTSilenceFlag is false
CM Server-MCS7800 - System performance and functionality degrades and I am having to reboot 1x/wk...memory leak ??? Do you have a reference to that fix ??
Thanks,
Rick
09-25-2001 07:25 PM
We had a simular issue, and we traced it back to VAD, even if you turn off VAD on the router it still is on for incoming calls. This can be tested by using the "show voice call summary" and watching an incoming call you will see that VAD is still on. To remove VAD off the incoming Interface you need to add the command""inbound called-number ." (dot) to the incoming dial-peer.
11-02-2001 02:36 PM
We also solved this problem with VAD.
09-25-2001 11:13 PM
Hi Markus,
I did face similar problem but it is pure Centralised IP telephony solution. What i have found is if you use Microsoft Event Viewer tool on Windows NT Server then you would find lots of "Warning" & "Error" messages. If you translate these messages on the Call Manager server then you could pin point the actual cause of your problem. I am doing it these days and i find it very useful tool to troubleshoot the problem.
By the way i have some QoS issues these days on the net.
Discover and save your favorite ideas. Come back to expert answers, step-by-step guides, recent topics, and more.
New here? Get started with these tips. How to use Community New member guide