[OpenSIPS-Users] high-availability - senario

[Julien Chavanton]

I still have a problem with binding all IP some headers do not have the IP set :
 
Record-Route: <sip:0.0.0.0;lr;>.
Via: SIP/2.0/UDP 0.0.0.0;branch=z9hG4bKfc58.1427b764.0.
 
Is there a way to tell Opensip to use the routing source IP, or other alternative ?
 
 

________________________________

From: users-bounces at lists.opensips.org on behalf of Julien Chavanton
Sent: Tue 02/02/2010 1:44 PM
To: OpenSIPS users mailling list; OpenSIPS users mailling list; OpenSIPS users mailling list
Subject: Re: [OpenSIPS-Users] high-availability - senario


This was well documented
 
in opensips.cfg
 
    listen=udp:0.0.0.0:5060
 
or for both TCP/UDP
 
    listen=0.0.0.0:5060

________________________________

From: users-bounces at lists.opensips.org on behalf of Julien Chavanton
Sent: Tue 02/02/2010 12:47 PM
To: OpenSIPS users mailling list; OpenSIPS users mailling list
Subject: Re: [OpenSIPS-Users] high-availability - senario


I deceided not to use a loadbalancer because, the IP address for outbound call would have been different and we want it to be completely transparent for existing Interco.
 
I went with heartbeat + mon 
I created sip.monitor
 
When the fail over take effect it is not listening on the virtual IP, how do I configure opensips to bind to 0.0.0.0:5060 ?

________________________________

From: users-bounces at lists.opensips.org on behalf of Nigel Daniels
Sent: Tue 26/01/2010 11:45 PM
To: OpenSIPS users mailling list
Subject: Re: [OpenSIPS-Users] high-availability - senario



have you considerd using keepalived with vrrp instead ?


On Tue, Jan 26, 2010 at 7:03 AM, Julien Chavanton <jc at atlastelecom.com> wrote:


	I am configuring high-availability with heart-beat + LVS + ldirector
	 
	I do not want to load balance but mostly make sure it will fail over quickly, I have found situation where the in LVS (UDP/TCP) connection never timeout when the remote IP send periodic OPTIONS request for example.
	 
	I beleive I will have to set very low UDP/TCP time-out, however  with such a low time-out I can not load-balance so I will use weighted Round-Robin with very high priority 65535 on the active one and weith 1 on the passive server.
	 
	Any other suggested way to cluster without load-balancing ?
	 

	_______________________________________________
	Users mailing list
	Users at lists.opensips.org
	http://lists.opensips.org/cgi-bin/mailman/listinfo/users
	
	




-- 
Nigel Daniels
Network & Systems Administrator 
ConnectAndSell inc.
(650)-533-2542 


-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.opensips.org/pipermail/users/attachments/20100202/3e2fe27e/attachment-0001.htm