Large CPU System Utilization and Very Poor Call Quality on Mitel 5000 System

[sgates]

We had a brand new Mitel 5000/250 controller installed in February. We have had many issues, and have implemented VLAN and QoS on our network (Cisco Meraki MS350 and MS250 switches that all the phones are plugged into), as well as upgraded the software, and did a direct connection between the controller and our main switch. We did this as users were randomly being disconnected from calls, and the call quality was so bad that our users have resorted to using cell phones instead of their desk phones. We are school district and have occasion to dial 911, this is an issue due to the call quality. This morning our system went totally crazy, and our PA system that is connected to the phone system was even as well. We've noticed that the issues seem to be associated with 100% CPU utilization and the large percentage of that being application instead of users. We've reached out to our phone installer and they keep pointing to our network. Is there a way in the Mitel controller to figure out what is causing the issue? I looked at the ipra log this morning and noticed lots of error 110 as well as some other issues. We are desperate to try anything to resolve this issue as it is impacting our school greatly. I've tried contacting Mitel, but they won't talk to me as I'm not a Mitel tech. We selected Mitel for its features, but have come to regret that decision with this latest major issue. Any suggestions would be greatly appreciated.

[Tech Electronics]

sgates,

Sorry to hear you are having so many issues with your new system. There are many reasons that the CPU Utilization could be 100% and one of them is even the reporting services that tell you about it; go figure. Anyway, if you really need some help with it then we would need to get some more information.

Do you have 1, 2, or 3 IP Addresses associated with your controller. That will tell us whether or not you have a Processor Expansion Card [PEC] and/or a Processing Server [PS-1]. The PEC adds resources and the PS-1 adds processing power. Also with the PS-1 you can get more information from the logs associated with it.

Now, typically what you would want to log at is whether or not this problem happens all the time or at certain times of the day. Do you have ports opened up to the outside world or to other networks that you don't completely control. Is there a lot of Unified Messaging boxes 100+ or are they set to use .mp3 instead of .wav files; which uses more processing power to convert it to .mp3 files.

Thanks,

TE

[acejavelin]

Some more specifics about your site would be helpful... Approximate number of phones, types of trunks, etc.

Do you have remote access to this site via port forwarding? We have seen tons of issues with ssh (port 22/tcp) being forwarded to the controller and hackers hitting it so much they actually shut down the system. Just throwing that out there at random, ran into it again recently.

[BlackSunshine]

What kind of trunks? Are the pc and phones on same network? If so have you killed out the data side to see if calls improve.  Have you wiresharked to see what else might be clogging network?  Have you verified there isn’t duplicate Ip addresses?  Can you troubleshoot after hours when network has died down?  The first network I cut over using meraki switches the IT guys had them set up wrong and we had same issue. Also the meraki switches have good diagnostic.