Page 1 of 1

WaitforSilence is killing the CLI interface ..

PostPosted: Mon Oct 13, 2008 11:57 am
by Nortelguy
All I get is "Got 0ms silence< 1200ms required" scrolling waiting for silence. Multiply this by the amt of calls that are actively waiting for silence and its a diaster. There has to be a way to supress this so the CLI is usable again .. any ideas ? :roll:

PostPosted: Mon Oct 13, 2008 6:27 pm
by mflorell
you can set your verbosity to 1 I believe

We are testing a new version of waitforsilence. Hopefully we'll be able to release it soon.

PostPosted: Mon Oct 13, 2008 8:54 pm
by Nortelguy
Ah, would be nice to keep the Verbosity set to 21, maybe in the new version of WaitforSilence you can define its Verbosity ? Also with the dialer not placing any calls the Waitforsilence inthe CLI never stops ..Ever, not w/o a reboot or service refresh.

PostPosted: Mon Oct 13, 2008 9:39 pm
by Nortelguy
touching more on the issue with waitforsilence never stopping. I placed a test call to my cell and when the script starts to listen for silence I hangup. The script never stops

PostPosted: Tue Oct 14, 2008 6:01 am
by mflorell
mcargile just posted a new waitforsilence on our download site last night:
http://downloads.vicidial.com/asterisk- ... rsilence.c

Try downloading that and recompiling Asterisk and see if that works better for you.

PostPosted: Tue Oct 14, 2008 10:03 am
by Nortelguy
I had the wonderful experience of not having my slack box booting up this morning. I called Mike for a quick tip and he suggested the logs were full. Sure enough I had 400 gigs worth of logs .. yay . The Kernel didnt like crashing so I had to copy a new one over. Applying this new c file ASAP!! Thanks again guys.

PostPosted: Wed Oct 15, 2008 11:50 am
by mcargile
I just made an update to app_waitforsilence.c on 08-15-2008 which should fix the hangup issue completely. http://downloads.vicidial.com/asterisk- ... rsilence.c

The original author was under the impression that the ast_waitfor() function would throw an error if the channel hung up, which it does not. I put a call into the ast_check_hungup() function and exit if it returns true, This fixed the problem.

Thanks for pointing this out. I was actually under the impression that the call to ast_waitfor() would actually throw an error on a hungup channel. Though I based this on the concept that the author actually knew what he was doing. It might have actually done so originally for all I know. I was actually thinking that the calls being stuck were caused by calls to IVRs where there was too much back ground noise.