After running Voicemail Pro on a Windows 2003 domain controller (running Microsoft DNS) fine for 6 months, Voicemail Pro decided to stop communicating with the IP Office after a reboot. It looked like everything was fine as I could load the Voicemail Pro admin interface, but obviously there was no communication between the IP Office and Voicemail Pro. This is because MS DNS decided to eat the ports needed by the Voicemail Pro to communicate with the IP Office (UDP 50791-50801)…If you experience Voicemail Pro Not Operational, but it seems fine… and you get this in Monitor:

3233269mS CMTARGET: 5.20.1 27 Q931 Trunk:5 CHAN=1: TimerExpired cause=CMTCNoAnswerTimeout
3233269mS CMTARGET: 5.20.1 27 Q931 Trunk:5 CHAN=1: Retarget NOANSWER EXCEPTED=00000001 ValidTargets=1
3233269mS CMTARGET: 5.20.1 27 Q931 Trunk:5 CHAN=1: Retarget on target_cfg_user=SMConference
3233269mS CMTARGET: 5.20.1 27 Q931 Trunk:5 CHAN=1: ADD USER: SMConference depth=1 disallow_cw=0 dnd=0 real_call=1 group_call=0 type(CMNTypeUnknown) incl(0x1) excpt(0x1), allow_redir(1) remote=00000000
3233270mS CMTARGET: 5.20.1 27 Q931 Trunk:5 CHAN=1: SELECT: TRY VOICEMAIL orig_hg() orig_user(8509)
3233270mS CMTARGET: 5.20.1 27 Q931 Trunk:5 CHAN=1: ADD VM TARGET
3233270mS CMTARGET: Voicemail Pro not operational
3233270mS CMTARGET: 5.20.1 27 Q931 Trunk:5 CHAN=1: ADD VM TARGET: FAILED availability=0
3233271mS CMTARGET: 5.20.1 27 Q931 Trunk:5 CHAN=1: VM targeting failed. Remaining on final target SMConference
3233271mS CMTARGET: 5.20.1 27 Q931 Trunk:5 CHAN=1: GetNoAnswerTimer:22

Microsoft DNS will fight for UDP ports with IP Office. One way to test this is to shutdown the MS DNS server, restart VM Pro and then restart MS DNS. If THEN your server is working you need to reserve the IP office ports so Microsoft DNS can’t grab them:


Instructions for 2003 and 2003 R2 server:

Start Registry Editor (Regedit.exe).
Locate and then click the following registry subkey:
HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\Tcpip\Parameters

On the Edit menu, point to New, and then click Multi-string Value. There may be an entry already there, just modify it.

Right-click the new value, click Rename, type ReservedPorts, and then press ENTER.

Double-click the ReservedPorts value, type the range of ports that you want to reserve, and then click OK.

Add 50790-50801

Click OK.

Quit Registry Editor.

Restart your voicemail pro system (full reboot)… now your MS DNS will not try to grab ports that VM PRO needs to operate.


Instructions for 2008 and 2008 R2 server:

Server 2008 and 2008 R2 do not support the above registry entry. You will need to download the following hotfix from Microsoft, install it, reboot and then run the command below.

https://support.microsoft.com/kb/2665809/en-us/

Command prompt (elevated):
netsh int ipv4 Add excludedportrange protocol=udp startport=50790 numberofports=12 store=persistent

Reboot.


Instructions for 2012 and 2012 R2 server:

Server 2012 and 2012 R2 do not support the ReservedPorts registry entry. You will need to run the command below.

Command prompt (elevated):
netsh int ipv4 Add excludedportrange protocol=udp startport=50790 numberofports=12 store=persistent

Reboot.


This article is also posted on Tek-Tips IP Office FAQs, which was authored by me.

 


Tagged with:  

A client of mine has eight Verizon FIOS analog lines that are delivered to two Avaya IP Office analog ATM4 trunk cards. All of their calls are picked up/answered live, so no attendant pickup during working hours.  We deliver these analog calls to a hunt group to ring several users in the front office.

Because of a very specific issue with FIOS handling of caller ID in combination with IP office analog cards, the following occurs:

  • The analog port, set to Loop Start ICLID picks up the ring.
  • The call is delivered directly to a “main” hunt group.
  • If a member of the hunt group picks up the call on the second ring, they will hear a loud buzzing/flutter for 1 second.  It sounds like this (recorded from speakerphone).  The sound is inaudible to the remote caller, FYI.

There is no real “fix” per se without completely disabling caller id on the carrier or IPO side, but there is a workaround which requires Embedded Voicemail, Voicemail Lite, or Voicemail Pro by utilizing “announcements.”

  • Enable Announcements for your “main” hunt group (or extension if not using hunt groups), but don’t record one.
  • Set announcement wait time to 0
  • Set post announcement tone to “Ringing”
  • Turn off 2nd and repeat announcements.

Because the call is picked up by the attendant and then routed, the crackle usually heard by the person picking up the call is intercepted by the “fake” attendant announcement.  Note that this workaround will occupy a single VM channel for about .5 secs when the call is picked up.  Also, caller ID continues to work well with the fix above.

Yet another alternative is to disable caller ID at the Verizon side, or set your IP Office analog lines to “Loop Start” with no CID.  Of course this will disable caller ID which is usually a dealbreaker in this day and age.


Tagged with:  

I’ve been experimenting with an Avaya IP Office Small Office Edition (4T, 8A, 3 VOIP) in preparation for rolling a few out to clients in the next couple of months.  I finally got around to hooking up my Vonage V-Portal to the IPO and was excited to start testing.

To my chagrin, there was really serious local echo (hear myself talking) on my 4620 IP phone when using either of my two Vonage channels.  This ECHO issues was isolated to external calls on the Vonage device.  I spent 3 hours experimenting in the Avaya Manager on both the VoIP extension forms and the Trunk configurations trying to fix it to no avail.  Sometimes the echo would eventually cancel out if you were on the phone long enough, but there was definitely something very wrong with my configuration.

I was Googling and found a forum post mentioning something about the impedance on the line being a possible cause for Echo on analog lines in an IPO environment.  So I tried each of the Impedance settings and sure enough the Impedance setting “Alternative 2” fixed the echo 99%. Other’s mentioned this change fixing POTS lines too.  Who knows, maybe it applies to more VoIP ATAs as well.  Keep me posted in the comments.

UPDATE

I also found that disabling HANDSET/HEADSET AGC (Auto Gain Control in the system menu) helps eliminate  voice clipping in conjunction with the V-Portal.

UPDATE 2

I gave up trying to use a V-PORTAL with my IPO.  The call quality wasn’t very good, with lots of clipping.  I don’t know if this is a Vonage issue or something to do with IPO talking to Vonage.  I ended up moving my lines to a SIP trunk with Vitelity.  I will post a blog entry detailing how I got that working (easy) soon.


Tagged with:  

“Hello operator?  Get me Chelsea 602 please.”
  • Project:
    Convert 1A2 system to Avaya Partner ACS, preserving existing cable/wire.
  • Scope:
    3 CO Lines (2 Voice, 1 Fax), 1 Gate Line, 10 Extensions, Fax Machine, Answering Machine
  • Parts Used:
    Avaya Partner ACS R6 Module, 308EC Module, 2-Slot Carrier, (2) 66-Blocks with preinstalled jacks, (1) Standard 66 Block, Amphenol to RJ11 Converters, APC Back UPS 650, RCA 5-Disc CD Player

This project was for my Grandmother’s house.  She’s had her 1A2 system for 40+ years and up until recently it has given great service.  Line 3 on the 1A2 was giving serious audible static to the point of not being usable.  Instead of troubleshooting and fixing the old system, I figured it was more cost-effective to put in a used, but more modern system.  I chose the Avaya Partner ACS as the replacement system as it’s cheap, full-featured and I have experience with it (I’ve owned a few).

If you don’t know what an Electromechanical 1A2 system is, think of it as sort-of the OG of phone systems. It was the first prevalent multi-line phone system.  You can read more about the 1A2 at Wikipedia. Here’s a picture of a 1A2 phone:

See that crazy connector hanging off the right of it?  That’s the amphenol connector which you may remember from a Parallel printer’s connection to a computer before USB became popular.  Each 1A2 phone required a 25-pair cable to operate and terminated with an amphenol connector at the phone end.  For our project we simply disconnected the old phones and snapped on an amphenol to RJ11 adapter.  In some circumstances we had to chop the amphenol off and use a biscuit, but the majority used adapters.  The good news is when you are replacing a 1A2 is you’ll have LOTS of existing cable to deal with.

In this particular project there were two blocks that cross-connect to the KSU, both requiring retermination.  One of the blocks was in a central spot near the system in the basement (Block A) which was connected to the KSU via another 25 pair cable.  This is where the majority of extensions terminated and where Block B cross connected as well.  There was another block (Block B) in the attic powering the upstairs 3 extensions (this was a surprise we found late in the job).  My task was to reterminate the cross-connect blocks so that every phone would have it’s own dedicated 2 pairs.  In a 1A2 system each phone “shares” a series of pairs with the other phones… it’s complex to explain and Wikipedia does a better job of it.

My friend Walter (thanks Walt!) and I completed the job in about 11 hours on a Saturday.  It worked out great and my Grandmother loves her new phone system.  Check out the pics below of the before and after.  You can also read more at this thread at Sundance Communication’s phone system tech support forum where I asked pre-install questions to the experts, having never done something quite like this before.

Before & After Shots:

Response code is 404


Tagged with:  
Web Hosting