No Audio / general system slowness

[cholzhauer]

We're fighting an issue where random calls have no audio, or the phone keeps ringing after it's been answered.  This applies to both IC calls and external calls.  Our vendor has replaced all of the guts in the phone system and we've opened several tickets with Mitel, but they appear to be getting no where.

Because of the fact this is happening to IC calls, it can't be a carrier issue.

CPU usage and memory usage are normal

[acejavelin]

We're fighting an issue where random calls have no audio, or the phone keeps ringing after it's been answered.  This applies to both IC calls and external calls.  Our vendor has replaced all of the guts in the phone system and we've opened several tickets with Mitel, but they appear to be getting no where.

Because of the fact this is happening to IC calls, it can't be a carrier issue.

CPU usage and memory usage are normal

Is SSH (port 22) forwarded from the public internet, perhaps for remote programming access?

[NTEDave]

Faulty switches?

If the phones aren't answering or you see the digits appearing slowly when you dial out it points to a bandwidth issue, as you are having issues internally on the LAN phone to phone it points to a switch issue or Ethernet loop??

If you are getting no voice intermittently this could point to a throughput issue on your switches.

[acejavelin]

@NTEDave This would likely raise A032 - Insufficient Bandwidth alarms though, I don't see how it couldn't if it was this severe. Although there is a decent chance this is network related issue.

I have also seen this exact behavior when the system was attempted to be hacked on SSH, no alarms, no high CPU usage, nothing unusual with the exception of thousands of log entries showing invalid SSH logins (can't remember which log to check at the moment), but the system just drags along, everything is just slow.

Some other issues could be: SIP attacks, excessive voicemail to email syncing or recording of long conversations, flooding of ISDN errors on PRI, and excessive use of G.729 without an expansion card or PS1.

[cholzhauer]

SSH is blocked from the outside.  We have both PRI and SIP, but SIP accounts for 99% of incoming calls.

The switches are Procurve 5406's and the phones are on their own VLAN (same as system) with QoS enabled.  If I have bandwidth issues on these $5,000 switches, I have  a big problem.

We do record conversations, but that's done via Tracer.

What's G.729?  We do not have the PS1, but CPU usage is normally under 25% (I don't know if that's a good indicator or not)

We do get insufficient bandwidth warnings once in a while since these issues started, but that's not normally something we see.

If it helps, here's a list of ports open to the outside

tcp-udp/69, tcp/3998-3999, tcp/44000, tcp/5566, tcp/6800-6802, tcp/6880, udp/20001, udp/5004-5007, udp/6004-6261

[NTEDave]

Insufficient bandwidth alarm can be toggled off and frequently is if the system has remote phones on ADSL connections. A lot of the systems I look after with remote phones would have permanent A32s as the 250 has a bit of a hair trigger for raising these alarms even when the end user suffers absolutely no issues.

Can the issues be re-produced or is it absolutely random?

[cholzhauer]

I guess I should say those alarms happen on the affected phones, it's never a system wide alarm.  (We do have 10-15 external phones on varying connections)

Absolutely random.  We had a bunch of errors yesterday morning, but none yesterday afternoon.

[NTEDave]

Is it random which phone is effected as well? Or could you narrow it down?

[cholzhauer]

Again, random; all phones are affected