VM Pro hunt group AA ring-no-answer 1

[jtierney]

We have an 8 site SCN (mix of IPO SOE, 403's, 406's and 412 ALL @ 3.1 v56 and VM Pro 3.1 v18). Central site has PRI and VM Pro server (WIN2K3). Remotes connected via PtP fiber (all units connected by CAT5 patch to Transition Networks 10/100 fiber tranceivers set to 10/H to a central 3COM 16 port auto-negotiating switch. The VM PRo server is also connected to this as well as feed to the customer network. Narrowing down the problem has been a challenge, but here is our best guess as to the scenario.

BOTTOM LINE: When a call is placed to a Hunt Group and no one answers, the call goes to a VM Pro Group>Leave action (menu w/ choices of who to call via Transfer actions.

IF I select an option to transfer to a member of the hunt group (lets say x1100, member of HG 1990), the call rings-no-answer w/ no visible attempt to cover to VM in Call Status.

If x1100 is on SAC/DND the call goes to VM

If I select an option to transfer to a person NOT in the hunt group(locally or somewhere else in the SCN) , the call goes to VM normally.

COMPLICATING FACTORS:

This does NOT happen with hunt groups @ the central site (not a PRI issue as it happens internally as well).

This is NOT a result of the outside fiber links as I now have a 403 in the central server room connected by fiber tranceivers the same as the remote sites and the same scenarios occur).

Using an Assisted Transfer action results in silence until the call goes to VM via a leave mail action from the No Answer box of the Assisted transfer, usually cutting off some or all of the greeting.

This seems particularly bad since we went to 3.0+.

JT

 

[CarGoSki]

I wonder if you are running out of data channels at the main site. Perhaps the call is not collapsing(hairpinning) properly. You should try to capture monitor and vmpro traces.

 

[jtierney]

Absolutely. I have sent traces for several calls to our tech support provider.

 

[CarGoSki]

You should post them here is what I meant to type.

 

[jconrad0305]

are all of your sites running the same version of software?

 

[jtierney]

All running 3.1 v56 with 3.1 v18 VMPro

LAB TEST:

Also, I just set up a lab SCN with a 403 and a 406V2 running the same releases. The SAME thing happens, this time on the same LAN.

Call FROM the IP 403 where the VMPro server is configured (from x1004) TO a hunt group at the remote IP 406 (HG 3020)

I let the call ring at the assigned member (x3003) station then cover to the VM Pro GROUP Leave Mail action, an auto attendant w/ touch tone 1 & 2 (1= x3003 @ remote site, member of HG 3020 2= x1002 @ central site from where I am calling from).

Option 1 = transfer to x3003 rings-no-answer

Option 2 = transfer to x1002, rings then goes to vmail box

Hmmm, methinks I found a bug.

JT

 

[CarGoSki]

SCN bugs just like that have been around forever. But I wonder if you have also built short codes that point from one IPO to the other and not simply rely on IPO to sniff out the extensions for you.

Also I would dump VMPro assister transfer and instead use transfer. Particularly for SCN but also generally speaking.

 

[jtierney]

Build short codes for extensions, not just hunt groups?

The assisted transfer is pretty unreliable, anyway throughout the SCN. The transfer action is working great except for this snafu.

JT

 

[CarGoSki]

Yes I always build codes for the extensions in an SCN environment. I kinda got into the habit of it. Also explicit #UserName.Leave shortcodes could help.

I am building an SCN tonight to prepare for the IPO ACE test and I may just run your scenario. I was once told that one of my SCN customers transfers too much! Naturally these were all bona fide SCN bugs. Most of which were alleged to have been repaired by 3.1.48 and 3.0.6x.

Certainly if you can duplicate in a lab environment then you likely have sniffed out an SCN bug... perhaps.

The biggest problems of days gone by were VCM resources not collapsing into one resource and one way audio. The first would exhaust vcm resources rather quickly and the second was very annoying.

It would appear to me that the the vcm resources are scrambled again.

 

[CarGoSki]

Well I did my SCN and verified that it is broken as you have described particularly when using a transfer action. However if you use assisted transfer that routes to a leave mail action for no answer and busy then it works.

Grp.Leave Menu>x??? selector code>Assisted Transfer-$Key>No Answer/Busy>Leave Mail-$Key

Phantom extensions and extensions that do not have a phone plugged in work correctly and cover immediately.

BTW my lab is 3.1.48

Good luck.

 

[jtierney]

The issue has been escalated to AVAYA Tier III (waiting for callback this AM).

As far as short codes for extensions, would you build generically, such as a 2 site SCN, with 1xxx at one site and 2xxx at another connected by a VOIP line group 8

Would it look like this?
(from the location w/ 2xxx extensions)

SC: 1xxx
tel no:1N
line: 80
feature:dial 3k1

 

[CarGoSki]

It would look exactly like that except that I usually buck the trend and always use dial and never dial3k1. Personal preference.

 

[jtierney]

I'll test that out with my lab systems, thanks. As for dial3k1, for some reason, that has always had less problems for us.

Anyway . . .I am waiting on AVAYA to establish a webex for me (minutes away). They have a patch they'd like to try. Keeping fingers crossed.

JT

 

[jtierney]

GOOD NEWS FOR A CHANGE:

AVAYA sent me 3.1 v5601 .bins. Things have begun working on my lab SCN. I will be upgrading the client this evening and will keep you posted.

JT

 

[CarGoSki]

Thats good news. Who did you work with at T3?

 

[jtierney]

Mike W in Denver. In fact, we found something new in the process and it is currently being looked at by development. New post pending new info/testing.

Relates to transfers (again from Group>Leave menu actions to $KEY transfers from x??? touch tones.

JT

 

[mikecastdogg]

I am also having a problem with direct transfer over the scn. If the extension is not picked up it willring no answer and never get to VM. If I use assisted transfer the problem goes away. Any thoughts. Could you send me that .Bin file?