So let's start with what we did onsite:
-Ticket requested to troubleshoot international calling. Found out that problem was with all abound dialing, and it was sporatic. Later found out that it wasn't just dialing, but in progrees outbounds would also be dropped periodically.
-Windstream informed us router had been changed twice
-Checked all programming for the system as it relates to the PRI. Found PRI only configured for 19 channels instead of 22. Changed config, but should not cause the problem.
-Assisted Windstream tech with testing
-At the time of the testing with Windstream, we got zero calls to fail
-One user did have a call to fail, we were unable to duplicate.
-Windstream left site
-I then did about 120 test calls, and the problems would come in bunches. 16 good calls, 12 bad, 16 good, 12 bad, 19 good, 6 bad...then it became more random
-Called Windstream and asked them to monitor the PRI. Made 40+ test calls and could not get one to fail. Released Windstream.
-Isolated each channel on PRI and made 4-6 test calls from each channel. No failures on any test calls.
The site contact on the ticket suggested we replace the phone system. Also told me the problems started after the cutover to Windstream.
This is what I think:
-The phone system is has no mechanical problem.
-The Cisco IAD has no mechanical problem.
-The two units are not integrating well together.
-It seems the problem is timing(?). The Cisco Unit has trouble sending the proper timing to the Merlin Magix. The Magix is set to get the timing from the network(the Cisco). We could set the Magix to control the timing, but the issue will still remain. I checked the error logs, and there are not a significant amount frame slippage over the course of a week. However, during the time of failure, the frame slips do spike.
It is my opinion, and it is just an opinion, that if we replaced the router and phone system 100 times, this will still be a problem. I look forward to any ideas and suggestions that anyone has that may lead us to a solution.