I have an unusual issue that I'm hoping someone can shed some light on. We have a customer running an OS 7200 with analog trunks and IP phones. A couple of weeks ago, the receptionist started complaining that she was having trouble answering calls.

What happens is that a new call will come in and ring to a station group that includes her extension and one other (the other extension is normally unoccupied). When she tries to answer, the ringing stops, but she can't hear anything. She will hang up and, a few seconds later, the phone will start ringing again. She picks it up and is able to answer the original caller. The person calling in is unaware of any of this. They just hear normal ringing until the call is successfully answered.

This seems to be happening on maybe 10 to 20 percent of the inbound calls. Most of them work fine. Any thoughts?

What software version is your system on? What type of IP phones are they? Using OAS card I am assuming. Is the alarm history saying there is any issues in the system with IP duplication?

The system was running 4.60, I think, when the issue started. We have since updated everything to latest release (v4.98). Phones are a mix of 5210 and 3105. System has a single OAS card. There haven't been any recent alarms logged in the history.

Do the phones have DT buttons for all lines or do they just have call buttons?

The answer to that question is "Yes".
It's a combination of the two. The receptionist and a few others have DT buttons, others just have call buttons.

As it turns out, this was not a system issue at all, despite the service providers' insistence that it was. We were able to reproduce the issue with a test set connected directly to the CO lines, bypassing the system.

We still almost missed it, because dialing each line directly worked fine. What we found, however, is that calls are typically routed through a virtual number that forwards to the lead number on the CO lines. When dialing the virtual number instead of dialing the CO lines directly, the problem occurred on a large percentage of the calls.

We were finally able to convince the provider to take a look at it and they made a change to how the calls were being routed. That appears to have fixed it.