Summary:ASTERISK-01960: Get chan_sip.c : Failed to grab lock trying.... ---------after one SIP to SIP call
Reporter:idej (idej)Labels:
Date Opened:2004-07-06 22:56:28Date Closed:2011-06-07 14:00:48
Versions:Frequency of
Environment:Attachments:( 0) messages
Description:I'm using Asterisk as a SIP proxy and making a call from one X-PRO Lite SIP phone to another via Asterisk (version downloaded last week & built on RH9.0 ). SIP phones are on separate servers on same LAN segment. I can make the first call and answer but on answer, Asterisk enters an infinite loop with the above message i.e. Failed to grab lock..trying. Asterisk is in a livelock state & can't be used from then on. No more SIP messages are processed by Asterisk in this state. I have seenm this fault in Mantis but there was no clear resolution - migrating to RH8 is not an option.


See messages file for console output.
show version CVS-HEAD-06/24/04-09:24:18
Here is my sip.conf:

; SIP Configuration for Asterisk
; Syntax for specifying a SIP device in extensions.conf is
; SIP/devicename where devicename is defined in a section below.
; You may also use
; SIP/username@domain to call any SIP user on the Internet
; (Don't forget to enable DNS SRV records if you want to use this)
; If you define a SIP proxy as a peer below, you may call
; SIP/proxyhostname/user or SIP/user@proxyhostname
; where the proxyhostname is defined in a section below
; Useful CLI commands to check peers/users:
;   sip show peers Show all SIP peers (including friends)
;   sip show users Show all SIP users (including friends)
;   sip show registry Show status of hosts we register with
;   sip debug Show all SIP messages

context=default ; Default context for incoming calls
;realm=mydomain.tld ; Realm for digest authentication
; defaults to "asterisk"
; Realms MUST be globally unique according to RFC 3261
; Set this to your host name or domain name
port=5060 ; UDP Port to bind to (SIP standard port is 5060)
;bindaddr= ; IP address to bind to ( binds to all)



Comments:By: Mark Spencer (markster) 2004-07-07 08:47:45

Find me on IRC, either sooner or after July 11 at worst and I'll need to login to look on your system.

By: idej (idej) 2004-07-07 17:00:32

One thing I should add to this - I am running RH9 as a VMWare Virtual Machine as the guest O/S under Windows 2000 Host O/S. I would not expect this to have a bearing on the fault but it just might.


By: Brian West (bkw918) 2004-07-07 17:14:22

someone said you have to tell vmware not to use the windows paging file. Otherwise it goes to hell fast.

By: Malcolm Davenport (mdavenport) 2004-07-08 13:03:17

Can you confirm that you have this problem outside of VMWare?

By: idej (idej) 2004-07-11 17:30:16

Unfortunately, I don't currently have a machine runnning RH9 ouside of VMWare. I am looking in to getting one however. I am struggling to see how Asterisk would be upset by VMware so if anyone has any ideas...

By: Mark Spencer (markster) 2004-07-12 09:25:53

Nothing occurs to me at the moment, but the fact that this happens with every first/second call on your system and nobody else has reported the problem strongly suggests that something in your configuration (be it VMWare or some combination of options that you've specified) is tickling this bug into existance.

If we can eliminate VMware, that's the easiest, then I can ssh in and look at the rest.

By: Olle Johansson (oej) 2004-07-15 14:05:12

Housekeeping gently reminds that we need an answer on this report. We'll close it while waiting unless something happens soon...


By: idej (idej) 2004-07-15 16:45:47

I'm currently building a machine with RH Linux exclusively - I will update this bug as soon as I haave Asterisk installed & have it all tested.

By: Brian West (bkw918) 2004-07-16 11:03:15

This is vmware related.  It has to be.  I have many boxes doing sip without this.  If for some reason its still a problem when you upgrade please post again and we will reopen it.