[asterisk-bugs] [JIRA] (ASTERISK-26746) pjsip: Repeated segfaults with not responding nameserver

[Michael Maier (JIRA)]

     [ https://issues.asterisk.org/jira/browse/ASTERISK-26746?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Michael Maier updated ASTERISK-26746:
-------------------------------------

    Attachment: backtrace-2017-01-24T08.54.06+0100.txt
                backtrace-2017-01-24T08.51.21+0100.txt
                backtrace-2017-01-24T08.30.27+0100.txt

These are the backtraces of the segfaults. Unfortunately, there is no debug output, but some errors in logfile:

[2017-01-24 08:30:17] ERROR[8143] res_pjsip.c: Error 320046 'No working DNS nameserver (PJLIB_UTIL_EDNSNOWORKINGNS)' sending OPTIONS request to endpoint telekomPJSIP
[2017-01-24 08:30:17] ERROR[8143] astobj2.c: FRACK!, Failed assertion bad magic number 0x0 for object 0x7f6214011260 (0)

[2017-01-24 08:51:13] ERROR[2255] res_pjsip.c: Error 320052 'DNS "Server failure" (PJLIB_UTIL_EDNS_SERVFAIL)' sending OPTIONS request to endpoint telekomPJSIP-xxx
[2017-01-24 08:51:13] ERROR[2255] astobj2.c: FRACK!, Failed assertion bad magic number 0x0 for object 0x7f267400a4b0 (0)

[2017-01-24 08:53:58] ERROR[4150] res_pjsip.c: Error 320052 'DNS "Server failure" (PJLIB_UTIL_EDNS_SERVFAIL)' sending OPTIONS request to endpoint easybellPJSIP
[2017-01-24 08:53:58] ERROR[4150] astobj2.c: FRACK!, Failed assertion bad magic number 0x0 for object 0x7fe0b800b3f0 (0)

easybell uses SRV lookups, telekom classical lookups.

> pjsip: Repeated segfaults with not responding nameserver
> --------------------------------------------------------
>
>                 Key: ASTERISK-26746
>                 URL: https://issues.asterisk.org/jira/browse/ASTERISK-26746
>             Project: Asterisk
>          Issue Type: Bug
>      Security Level: None
>          Components: pjproject/pjsip
>    Affects Versions: 13.13.1
>         Environment: Centos 6
> dual core processor
> FreePBX 13.0.190.7
>            Reporter: Michael Maier
>         Attachments: backtrace-2017-01-24T08.30.27+0100.txt, backtrace-2017-01-24T08.51.21+0100.txt, backtrace-2017-01-24T08.54.06+0100.txt
>
>
> Asterisk was started while DNS was not functional: local named (configured in /etc/resolv.conf) was reachable, but named couldn't resolve any hostnames, because of broken INPUT iptables rules (the answer of the provider DNS was never seen by named). Therefore request timed out.
> There have been trunks configured using SRV and not SRV entries. I can't say, which of both types (or even both) caused the segfaults. Unfortunately, no debug was activated at the time the crashes appeared.



--
This message was sent by Atlassian JIRA
(v6.2#6252)