the key to this type of configuration is starting the openser process on the machine that takes over the ip, AFTER it has taken over the ip, or it will not be available to openser at start time. keepalive allows for some simple script execution at times of failover to slave, and reacquition of the ip by the master when it comes back up

On Tue, Nov 11, 2008 at 11:57 AM, sergejf <serge@elasticall.com> wrote:

We would like to set up two OpenSER 1.4.x servers in a failover
configuration. The active (primary) server has two main roles: 1) load
balancing SIP INVITEs between SBCs from various carriers using the
CarrierRoute module 2) collecting CDRs from all these PSTN calls using the
Acc module and storing them in a MySQL database for later billing
reconciliation.

Right now we're using a F5 Big-IP load balancer to handle the pool of two
SIP proxies which receives the SIP INVITEs from our Media Servers and
dispatches them to the SBCs. However, should the primary proxy fail, the
secondary proxy would never receive (for example) the SIP BYE from an SBC.

Solving that issue with the Big-IP seems complicated and a simpler
vrrp/ucarp solution looks attractive. Any (hands-on) advice on how to set up
vrrp/ucarp to advertise a single virtual IP address, then detect that the
active SIP proxy has gone down and switch to the secondary IP, would be most
welcome. We're using Centos 5.x (64 bit) on our servers.

Thanks a lot.

Serge


--
View this message in context: http://www.nabble.com/OpenSER-HA-example-with-vrrp-ucarp--tp20443679p20443679.html
Sent from the OpenSER Users Mailing List mailing list archive at Nabble.com.


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