Recently we faced an issue with our Mitel IP telephony system after upgrading from version 7 to 8, then to 9.

We’ve Nortel (Currently Avaya) legacy digital telephony system across our businesses and the Mitel IP telephony was implemented with an intention to roll out IP based telephones. Somehow, the solution was partially implemented, forcing us to integrate two different technologies, so that we can make and receive calls from and to both systems.

The digital trunk was created using E1 interface card and everything was working until the software upgrade. We started experiencing a kind of annoying noise, almost impossible to hear the other party & our support partner from Mitel spent 3 weeks trying to figure out what went wrong after the software upgrade.

After 3 weeks of involving different “engineers”, they gave us 3 different possibilities for the noise:

• Controller’s motherboard have issues
• Controller’s hard disk is having issues
• LAN is having bad POE switches

We’ve downgraded from version 9 to 8, “successfully” inheriting the noise issues & and we had a new problem in addition to the noise issues. Handsets started reloading the application once in 2-3 minutes. This is the time we (internal IT team) decided to intervene.

We’ve identified the application reload was happening due to a power failure, which resulted in the IP address used by the Layer 2 (L2) Switch being issued to one of the PCs in the network as it was not excluded with DHCP scope. After spending couple of hours being assigned to address this situation, support partner concluded that “it looks like a hard disk failure”

We took over & started checking for the IPs assigned to the controller and found that it was merely an IP conflict and fixed the reloading issues by releasing the IP (added an exclusion scope with the DHCP server to avoid future conflicts)

However, we had the bigger issue still not resolved, about the noise. As we realized that the support partner is JUST making vague and unrealistic conclusions, requested for the details about codecs and the settings for trunks, which they happily shared with us.

By default, the codec settings for our controller running version 8 has values as seen with the below image.

We toggled all the switches (allowed ones) to No first (Saved), then Yes (Saved) followed by the default settings (refer the image)

Went ahead to “Trunks -> Digital -> Digital Link Descriptors” & re-assigned Voice Encoding to “ADI” (Which the “engineers” set as Nil during their extended support efforts)

Saved the changes and 3 weeks long noise issues were gone in a second.

Although we can’t confirm (technically), it looks like the backup restore after the software upgrade somehow interferes with the codec filtering that needed a manual intervention like the one we did by toggling the switches to NO and Yes & defaults under Codecs section.

If you were frantically searching for solution such a situation like ours, before falling victim to your support partner’s offer to replace the current controller for a discounted price like 4000$, please spend a moment to give what we have tried and succeeded a try.

regards,