I just ran into an issue that I have not seen before. An IOS voice gateway running 12.4(11)T4 has a dot1q trunk with several logical interfaces off of one physical. MGCP was using one of these logical interfaces to bind control and media. All was well, and the interface was up.
I cannot find what caused this, but all of a sudden the voice gateway becomes unregistered, and a 'sh mgcp' shows "MGCP Admin State ACTIVE, Oper State DOWN - Cause Code IP_INTERFACE_NOT_UP". The interface was up, and never went down. I could do an extended ping to the callmanager using this interface as a source, and was successful.
In a pinch, I removed the mgcp binding, and the gateway instantly registered.
In a maintenance window, I am going to create a loopback on the router, and bind to it. The question is, why would this happen, and has anyone else seen a similar problem? I have searched the bug toolkit, and am coming up blank.
I got to work this morning and got that bug-id for ya!
It's bug id CSCso53115
Like i said, I saw the same symptoms but a no mgcp/mgcp brought everything back online.
The symptoms of that bug say its partially reproducible. Well every time we bounced this router it happened. Regardless of what we tried to change in this configuration. TAC had told me that this was a known bug in earlier 12.4 versions. Dont have a ID on those bugs though.