SIP Failover

[billmatthews]

We have a two data center environment (primary and DR).  Each site has a CUCM subscriber, a cube router with a SIP to Verizon, and a PRI router.  Primary site has the pub.  This environment is 100% managed by a managed service provider, so my questions are just for my education.  I don't need any config statements, just high level understanding of how it should work.

For inbound calling, I think all calls come into the primary cube, and they use rollover (down or busy) to the DR cube.  Is that the standard design for inbound sip failover?  Any other issues we might want to consider?  How would it work with multivendor providers (not both Verizon)?

For outbound calling, how do we failover from primary_cube_sip, to dr_cube_sip, to pri as a last resort?  Is that just done through route groups/lists?

If there's some document that might help me understand this, I'm more than happy to read.  But most of the Cisco technical docs have been too deep for me at the moment.

Thanks

Bill

[Ayodeji Okanlawon · Accepted solution]

Bill,

The DR/fail over is done like this..

1. On the outbound direction, you use a combination of Route List, Route groups and OPTIONS ping to detect the failure of one of the CUBEs and the call is routed by cucm to the second CUBE

2. On the inbound direction, there are two scenarios to consider

a. If the ITSP (Verizon) cant reach your primary CUBE, then they swill send the call to the secondary CUBE. This is done at their side and there is nothing you can do about that. Infact in our case, Verizon sends the call in a round robin fashion to both of our CUBES to distribue the load on the CUBES..

b. If the call arrives on the CUBE and the CUBE cant reach the primary cucm, a combination of OPTIONS PING, SIP timers and dial-peers are used to send this to another CUCM...

Please rate all useful posts

"opportunity is a haughty goddess who waste no time with those who are unprepared"

[billmatthews]

Thanks -- very helpful!