Solved: Call server Fail over

abraham23482 · ‎12-12-2009

Dear All,

We have a distributed setup of Contact center with 2 call servers in 2 different locations. Dial-peer configuration in site-1 has been made in such a way that if the call server in site-1 goes down calls should be directed to call server in site-2 by means of preference statement in dial peer.

But it is noticed that when the call server at site-1 is down and when we make test calls at that time, there is a huge delay get the call answered.

Please suggest the possibilities of this problem.

geoff · ‎12-12-2009

What are you specifying in the dial peer? You should refer to the SIP proxy server, not the Call Server.

(Maybe you don't have a pair of CUPS - not the way I would do it, although Cisco say this is allowed; definitely inferior.)

You idea is workable in the absence of a SIP Proxy, though. The first preference dial peer points at one CVP and the second preference dial peer points at the other CVP. The SIP user agent on the gateway is configured with a default set of retries on the SIP INVITE, and a delay between retries.

When the first Call Server cannot be accessed, the gateway will not drop down to the second preference dial peer until it's retried the first one. You need to wind down the number of retries and the delay between retries. The default on retries is 6.

Do show sip-ua timers. The normal setting on the INVITE is 500ms. So it will take three seconds before the gateway gives up on the first dial peer.

In the SIP section of the gateway:

sip-ua

retry invite 1

Regards,

Geoff

View solution in original post

geoff · ‎12-12-2009

What are you specifying in the dial peer? You should refer to the SIP proxy server, not the Call Server.

(Maybe you don't have a pair of CUPS - not the way I would do it, although Cisco say this is allowed; definitely inferior.)

You idea is workable in the absence of a SIP Proxy, though. The first preference dial peer points at one CVP and the second preference dial peer points at the other CVP. The SIP user agent on the gateway is configured with a default set of retries on the SIP INVITE, and a delay between retries.

When the first Call Server cannot be accessed, the gateway will not drop down to the second preference dial peer until it's retried the first one. You need to wind down the number of retries and the delay between retries. The default on retries is 6.

Do show sip-ua timers. The normal setting on the INVITE is 500ms. So it will take three seconds before the gateway gives up on the first dial peer.

In the SIP section of the gateway:

sip-ua

retry invite 1

Regards,

Geoff

abraham23482 · ‎12-12-2009

Thanks Geoff.

We do not have a SIP Proxy. I check the timer and it was the default value. I am waiting for an approval from Mgmt to reduce this.

Thanks again and Merry Christmas

geoff · ‎12-13-2009

Don't alter the timer - just set the number of retries to 1 for the SIP INVITE as noted above.

Regards,

Geoff