Cisco Support Community
Showing results for 
Search instead for 
Did you mean: 

Welcome to Cisco Support Community. We would love to have your feedback.

For an introduction to the new site, click here. And see here for current known issues.

New Member

Failover for SIP CUBE


I have a question of how to maka a failover between two CUBE's placed in two different sites with no layer 2 routing and different IP networks. Is it possible to let the CUCM 6.2 ti fix this with two SIP trunks configured and one is primary or do you need HSRP? BGP is not an option! Regards Stellan

Everyone's tags (2)
Cisco Employee

Re: Failover for SIP CUBE

Can you purchase separate SIP trunks for each CUBE?  Or at least have the provider accept registration on the second as a 'failover' trunk?  Because if you can, 15.1(2)T has mechanisms built into CUBE for box-to-box failover with HSRP.

Check out this link for more details:

New Member

Re: Failover for SIP CUBE

Hi Steven,

we have to SIP trunks from a vendor located on different sites (IP/VPN routers) and there is no possibillity to use a Layer 2 routing between the sites so the Lan addresses are not the same so HSRP can't be used. The question is: Can Call Manager, CUCM, handle the failover with one SIP trunk to each CUBE configured?

Cisco Employee

Re: Failover for SIP CUBE

I assume you are using CUBE in media flow-through mode  If so, if you aren't going to use HSRP, then you'll never get the calls to be preserved during an outage.  However, new calls could be sent to the backup CUBE simply by having the provider configured to send calls to the other CUBE IP when the first is down.  And on the CM side, you'd have another RG in the RL for the backup CUBE, so that outbound calls hunt to that guy when the first is unresponsive.

If you did flow-around mode, media wouldn't go through CUBE, so you could achieve call survivability on active calls if your provider disabled session refresh for the trunks.  Because CUBE is stateful, when the original CUBE goes down, call control goes away and session refresh wouldn't be able to be sent from the backup CUBE.  Even the box-to-box redundancy feature on CUBE with HSRP can't handle session refresh, though, since the failover isn't stateful (yet).