There are no answer for the SIP invite on cisco gateway

[Mikhail Lapin]

Hello for everybody

I have CUCM 11.5 and two voice gateways connecting to PSTN ISR4431 and 3925E. For the outbound call I use Route List which contains two Route groups. Each Route Group contains SIP trunk to voice gateway. Today I tried to check redundancy and shutdown interface to the PSTN on the first gateway (4431) in Route List. Then I look debug ccsip mess on both gateways. CUCM sent Invite to the first gateway (4431) with shutdowned interface and after that gateway tried to sent more than 10 Invites although the interface is still shutdown and only after 2 minutes gateway sent "SIP/2.0 408 Request Timeout". Then CUCM took second trunk (3925E).

What can I do in this situation? How can I improve it?

Thank you

[R0g22]

Try reducing the "retry invite" timer under "sip-ua" and see if that helps.
I am confused understanding the fact where is the GW sending an invite ? Is this CUBE ? Which interface did you shut down ?

[Mohammed al Baqari · Accepted solution]

Apply the following configs on your routers:

sip-ua
retry invite 3
timer trying 200

This will reduce failover time to 0.6 sec. Also seems that you are using
CUBE, I suggest to implement TCP SIP instead of UDP sip. This will be
faster as gateway will fail to establish TCP session and hence no INVITE
sent. For UDP it will send INVITE even if interface is down.

Last implement options-ping on CUBE to monitor the provider and speed the
failover. In this case CUBE will busy-out the dialpeer to provider if it
fail to response and when it receive message from CUCM, it will immediately
respond with error.

[Mikhail Lapin]

Mohammed thank you, it works faster, delay becomes to 7-8 seconds. Could you suggest how can I use TCP SIP? Which commands I sould use?

The suggest with option-ping is very interasting too. Do you have configutation example.

Thank you again

[Randall White]

You can configure Options Ping on the CUCM.

Go to Device -> Trunk and find the SIP trunk that points to your gateway (CUBE).

Scroll down to the bottom and find the name of the SIP Profile on the Trunk.

Now go to Device ->Device Settings ->SIP Profile and find that same SIP Profile.

Scroll down to the bottom, and check the box named "Enable OPTIONS Ping to monitor destination status..."

You will need to Save and Reset the trunk, do this after hours.

After a few minutes, go back to Device->Trunk. The page should now show status "Full Service" and time since the last reset.

[Mikhail Lapin]

Ok, I understood, this option already enabled.

[Mohammed al Baqari]

Hi,

TCP instead of UDP should be agreed with your provider. It can't be enabled
one sided.

Option-Ping on CUBE:

Here you go
https://www.cisco.com/c/en/us/td/docs/ios/voice/cube/configuration/guide/vb_book/vb_book/vb_9321.html

This will definitely improve performance as it marks dialpeer dead when
reachability is down

[Dennis Mink]

Have you turned on OOB options ping on each trunk, so cucm can flag it as down and go to the next trunk in the RG?