Hi Graziano, Thanks for testing, unfortunately Asterisk does not always return with an error on a segfault, so we have to also look back in the logs for the segfault.
Back to the basic issue, is there any reproducible pattern that causes the segfault? Are you using DAHDI or mISDN for your B400M isdn bri card ? Is it always during a call (when it segfaults) ? What is typical call volume? We may not have an answer for you, but this kind of documentation is useful. Lonnie On Sep 14, 2011, at 7:46 AM, Graziano Brioschi wrote: > Hi Lonnie > > I can confirm you that the "tail -n3" change in the "safe_asterisk" command > catch an asterisk crash and restart it...sob sob... > > Here a sample of email message generated by safe_asterisk: > > Asterisk on 'pbx-lecco' exited on signal 99. Restarting Asterisk... > > [Generated at 11:35:19 on September 14, 2011] > > > bye > Graziano > > > Il 08/09/2011 10.50, Lonnie Abelbeck ha scritto: >> Hi Graziano, >> >> I already changed the SVN so AstLinux 0.7.10 will use "tail -n3", it should >> work for you, so you might test to be sure. (It is unusual the mostly empty >> "user.info kernel:" line is added.) >> >> I agree "-n4" or "-n5" probably should be OK, but if "-n3" works it is best >> to minimize the look-back window. >> >> Or, I could be persuaded to use "tail -n5" if you can convince me. :-) >> >> Lonnie >> >> >> On Sep 8, 2011, at 2:24 AM, Graziano Brioschi wrote: >> >>> Thanks Lonnie >>> >>> yesterday morning I already have replaced the "tail -n2" command with "tail >>> -n5" in then "safe_asterisk" command; do you prefer a "-n3" or i can leave >>> the "-n5" option? >>> >>> Graziano >>> >>> >>> Il 07/09/2011 18.02, Lonnie Abelbeck ha scritto: >>>> Hi Graziano, >>>> >>>> I have not seen such asterisk 1.4.42 segfaults, but to help Safe Asterisk >>>> to catch the segfault (and auto restart asterisk) you can try this change: >>>> -- >>>> $ service asterisk stop >>>> >>>> Edit "/usr/sbin/safe_asterisk"... >>>> >>>> Search for the string "tail -n2" >>>> >>>> Replace with "tail -n3", save change. >>>> >>>> $ service asterisk start >>>> -- >>>> >>>> Let me know if this automatically restarts Asterisk when a segfault occurs. >>>> >>>> Lonnie >>>> >>>> PS: Before upgrading in the future be certain you first undo the change >>>> with: >>>> -- >>>> $ service asterisk stop >>>> $ rm /oldroot/mnt/asturw/usr/sbin/safe_asterisk >>>> $ service asterisk start >>>> -- >>>> >>>> >>>> On Sep 7, 2011, at 7:46 AM, Graziano Brioschi wrote: >>>> >>>>> hi list, >>>>> >>>>> I have a crash problem on a astlinux 0.7.9 system with asterisk 1.4.42. >>>>> Asterisk hangs randomly on this system with a messages like this (logged >>>>> in /var/log/messages): >>>>> >>>>> ..... >>>>> Sep 7 11:09:36 pbx user.info kernel: asterisk[4269]: segfault at 84 ip >>>>> 40012e2c sp bd1fcb7c error 4 >>>>> Sep 7 11:09:36 pbx user.info kernel: in >>>>> libpthread-0.9.28.so[4000d000+a000] >>>>> Sep 7 11:09:36 pbx user.info kernel: >>>>> Sep 7 11:09:37 pbx user.info safe_asterisk: Asterisk shutdown normally. >>>>> ....... >>>>> >>>>> Someone have some suggestions? >>>>> >>>>> The system is installed on a OpenVOX IPC110 embedded motherboard with >>>>> 2GB RAM, CF Sandisk UltraII 4Gb, one OpenVOX miniPCI B400M isdn bri card >>>>> and 20 Yealink SIP-T22P. >>>>> >>>>> Thanks >>>>> Graziano >>>>> >>>>> -- >>>>> >>>>> Graziano Brioschi ------------------------------------------------------------------------------ BlackBerry® DevCon Americas, Oct. 18-20, San Francisco, CA Learn about the latest advances in developing for the BlackBerry® mobile platform with sessions, labs & more. See new tools and technologies. Register for BlackBerry® DevCon today! http://p.sf.net/sfu/rim-devcon-copy1 _______________________________________________ Astlinux-users mailing list Astlinux-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/astlinux-users Donations to support AstLinux are graciously accepted via PayPal to pay...@krisk.org.
