Sorry send the message to the wrong address,so here again. ----- Hello again,
Q1: There is a little misunderstanding. The kernel doesn't crash also asterisk doesn't crash it only disconnects the parcipians. Most times it breaks up with following error message "P[Channel] no voice data" but i also had some occurances without a error message. I tried to debug the problem but i can't identify any errors in the protocol message or in the asterisk debug. It throws this error and the channel breaks up after this message. You can see the normal cleaning messages after this. Most times the phone on the digium card even rings correctly but if you take up the phone the channel breaks up. Q2: System: Pentium M 1.8Ghz, 1GB Memory, No Swap because system comes from SD-Card. Tmpfs file system for system logs etc. memstat of running system says: 150MB used up by system. 800MB free, about 100MB allocated for caches. Connections: asterisk pabx: 10 ISDN Phones on 10 Ports evenly divided on both beronet cards. 2 Phones via SIP-Ethernet. foreign reception pabx: 12 Reception Phones on a reception pabx (no asterisk system) connected via first E1-Port of the digium card. Versions: Asterisk 1.2.12.1, chan_misdn: 0.3.1-rc23 (hope it's the right one.) or state from 02.01.2007., Kernel 2.6.18.1, Zaptel: SVN-branch-1.2-r1317. (Were the only versions i found in my hurry :) ) Simple Test: Made multiple connections via normal phone calls. Monitor system ressource growth via ssh session and using a mixture of top, vmstat, iostat, memstat etc. After all mentioned phones were connected the system load was in the range of 0.2-0.4. Memory and Fileressource doesn't seem to be low. (In using this tools i can't be realy sure) (24 connections with 12 channels were established in this test) About 2 minutes after all phones were connected two voice-connections went down with the asterisk error "No space left on device". The zaptel side of these connections doesn't seem to be down and the phone on the foreign pabx stayed in connected state. After rethinking the happenings i'am not sure if the logical connection state of the misdn side was realy broken, i forgot to ask my college this..... i will rerun the test to check this the next days. ( The test should analyse system load without hardware bridged calls.) But anyway maybe there is a buffer setting i can increase ? g. Sven _______________________________________________ Misdn-asterisk mailing list [email protected] http://lists.beronet.com/cgi-bin/mailman/listinfo/misdn-asterisk
