Hello,
When my asterisk server reach 50 concurrent active calls, PDD increase from 450 to 30.000.
I have 6CPUs and 32GiB memory size on my debian 12.
There are limits in my asterisk configuration, that i must increase or any others things to check ?
Google fails to give a definition of PDD that would make sense here.
@david551 It’s probably post dial delay, the time between dialing and getting the first non-100 response essentially. I’m not sure what “increase from 450 to 30.000” means though.
A SIP log and console output would really be needed.
yes, it’s post dial delay. I tried to test my asterisk server capacity and untill I reach 50 concurrent active calls, my PDD values range from ~200ms to ~400 ms, after that PDD range increase drastically from 7000 to 40000 ms PDD/per call.
My target it’s to increase asterisk server limit to 200 concurrent active calls with PDD values similar to first 50. My asterisk server hardware have the following configuration: 6 CPUs and 32GiB memory space available. What I should check in this scenario ?
There is no answer to that, because it has to be understood where it is being introduced and you haven’t provided logs.
Note that, for anything timing related, you really need the actual log files, not the console output, as the latter doesn’t contain complete timing information. I’d also suggest increasing the resolution of the time stamps on the logs. (I think there now is an option to include more complete timestamping to console output.)
Also I would expect post dial delay to depend on Busy Hour Call Attempt type statistics, not on Erlang occupancy, unless, maybe there was so little memory that normal calls were causing thrashing.
This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.