Cisco Support Community
Community Member

MGCP Gateway on Loopback address

I was wondering of there is any reason not to use the loopback address on a 2821 gateway, I have 2 vwic2-mft-t1/e1 cards and callmanager 4.1.3.

There seems to be a debate on whether it is a good idea to use the loopback address or not.

Anyone have any thoughts on pros and cons


Re: MGCP Gateway on Loopback address

From a routing perspective, loopbacks are a great solution as far as reachability of a device is concerned. I have run into an issue where i had a loopback interface on the router, non routable. Had that interface on for testing. When MGCP process reset, it for some reason got bound to the loopback address and started causing oneway audio issues, as the loopback was non routable.

So definitely make sure that you use a routable loopback address and also use the bind commands for media and control to the loopback interface.

I have also seen another bug (dont recollect the bug id) where even after physically binding the loopback using the bind commands, when mgcp process resets, the bind media and bind control command seems to lose the ip address. (Sh mgcp shows blank output).



PS: please remember to rate posts!

Community Member

Re: MGCP Gateway on Loopback address

Thanks for the reply. My understanding of mgcp is that the bind commands default to the loopback address that is why I was considering this as a viable option.

In most cases I use a fe port however I have seen a very large deployment where due to issues started using theloopbacks and this has resolved their issues.

It certainly is an interesting topic as everyone has an opinion.

Thanks for your reply

CreatePlease to create content