Distributed VoiceMail on Application server 1

[bmast]

I have been troubleshooting for far too many days in order to get distributed voicemail on the application servers operational. The configuration process should be fairly simple, yet I don't seem to be able to get it working.

TEST CONFIGURATION
SYSTEM A
IP0 500V2 R8.0.42 "ipoanc.alcantellab.com" 192.168.101.20
VM PRO R8.0.1009 "vmproanc.alcantellab.com" 192.168.101.12

VM SMTP Configuration:
Mail Domain: vmproanc.alcantellab.com
Mail Server: vmproanc.alcantellab.com
Sender: vmproanc@vmproanc.alcantellab.com

IPO A Configuration
Voicemail = Centralized
IP Address = 192.168.101.12

SYSTEM B
IPO 500V2 R8.0.42 "ipofai.alcantellab.com" 192.168.125.20
VM PRO R8.0.1009 "vmprofai.alcantellab.com" 192.168.125.21

VM SMTP Configuration:
Mail Domain: vmprofai.alcantellab.com
Mail Server: vmprofai.alcantellab.com
Sender: vmprofai@vmprofai.alcantellab.com

IPO B Configuration
Voicemail = Distributed
Line = 17 (SCN fallback for IP Phones only)
IP address = 192.168.125.21

DNS Server (windows 2008) at 192.168.101.11
domain alcantellab.com
vmprofai 192.168.125.21
vmproanc 192.168.101.12

Testing Resuls
1) I can log in as root on both systems and ping the FQDN of the both servers.
2) I can telnet to both servers on port 25 and receive the SNMP welcome message.
3) Both servers and IPO have been rebooted multiple times.
4) Voicemail B says "distributed voicemail" when I log in using teh admin client.
5) Voicemail A does not display any 'distributed voicemail' servers in the list
6) Information does not replicate.

System B Webcontrol Error Log
Webcontrol on system B displays the my attempt to connect to SMTP sessions via telnet (I turned SMTP logging on), but shows errors the specific message:
1) "SyncStatus Slave: vmsyncmaster@vmproanc.alcantellab.com, vmproanc.alcantellab.com running ..., Distributed, Remote, Not Registered, TX(14/5/2012 15:31:57)RX(31/12/1900 16:00:00) Q(0)Unitiialized, Alive" - this repeats several times then...
2) "SyncTraceMsg slave vmprofai.alcantellab.com sending "Create: Master" to vmsyncmaster@vmproanc.alcantellab.com" then error appears
3) bool CSMTPConnection:Connect(SQLU ... ): Could not connect to the SMTP server on port 25 ... name or service unknown.

System A Webcontorl Erros
1) can't find correlating errors.

 

[tlpeter]

I had exactly the same testing results but then on two Windows 2008 servers.
I gave up and just use them in a SCN.
I have no time to raise this as a bug so i just continue.
Maybe you can raise a bug?

BAZINGA!

I'm not insane, my mother had me tested!

 

[bmast]

I have noticed that the response on the telnet to SMTP port 25 varies according to the release and how the FQDN is entered in the application server. I wonder if this is related? Specifically:
1) if the system is installed clean on the latest release of voicemail, and the FQDN is entered under hostname (vmproanc.alcantellab.com); then the /etc/hosts file will be built as follows:

127.0.0.1 localhost.localdomain localhost
::1 localhost6.localdomain6 localhost6
192.168.101.12 vmproanc.alcantellab.com vmproanc

2) As a result of the above, when you telnet to port 25, the welcome message displays "localhost.localdomain" instead of the name.

3) if you modify the name through the webcontrol interface and reboot; the /etc/hosts file is re-written as:

127.0.0.1 localhost.localdomain localhost vmproanc
::1 localhost6.localdomain6 localhost6
192.168.101.12 vmproanc.alcantellab.com vmproanc

with the vmproanc.alcantellab.com on the same line as the 127.0.0.1 entries instead. Because it is the last item on the line the SMTP HELO screen still shows localhost.localdomain (further modifications only change the first '127' line entry, appending at the end)

4) the dnsdomainname -f command displays localhost.localdomain Hostname = vmproanc I don't believe this adheres to best practices for smtp, so I am curious as to the design intent.

5) If you install the system and put only the Host portion (vmproanc)of the FQDN under hostname during setup, the install program instead modifies the /etc/hosts file to read:
127.0.0.1 vmproanc localhost.localdomain localhost
::1 localhost6.localdomain6 localhost6

as a result the HELO message shows 'vmproanc' instead of localhost.localdomain

 

[bmast]

After the above result I ran the test again entering only the host portion of the FQDNTEST CONFIGURATION

SYSTEM A
IP0 500V2 R8.0.42 "ipoanc.alcantellab.com" 192.168.101.20
VM PRO R8.0.1009 "vmproanc.alcantellab.com" 192.168.101.12
Entered only vmproanc under hostname during setup

VM SMTP Configuration:
Mail Domain: vmproanc.alcantellab.com
Mail Server: vmproanc
Sender: vmproanc@vmproanc.alcantellab.com

IPO A Configuration
Voicemail = Centralized
IP Address = 192.168.101.12

SYSTEM B
IPO 500V2 R8.0.42 "ipofai.alcantellab.com" 192.168.125.20
VM PRO R8.0.1009 "vmprofai.alcantellab.com" 192.168.125.21
entered only "vmprofai" during installation

VM SMTP Configuration:
Mail Domain: vmprofai.alcantellab.com
Mail Server: vmprofai
Sender: vmprofai@vmprofai.alcantellab.com

IPO B Configuration
Voicemail = Distributed
Line = 17 (SCN fallback for IP Phones only)
IP address = 192.168.125.21

DNS Server (windows 2008) at 192.168.101.11
domain alcantellab.com
vmprofai 192.168.125.21
vmproanc 192.168.101.12

Testing Resuls
1) I can log in as root on both systems and ping the FQDN of the both servers.
2) I can no longer ping only the hostname of the opposite server
3) The generated /etc/hosts file now reads:
127.0.0.1 vmproanc localhost.localdomain localhost
::1 localhost6.localdomain6 localhost6

4) I can telnet to both servers on port 25 and receive the SNMP welcome message with the hostname in the HELO message
5) Voicemail B says "distributed voicemail" when I log in using teh admin client.
6) Voicemail A does not display any 'distributed voicemail' servers in the list
7) Information does not replicate.

System B Webcontrol Error Log
Webcontrol on system B displays the my attempt to connect to SMTP sessions via telnet (I turned SMTP logging on), but shows errors the specific message:
1) "SyncStatus Slave: vmsyncmaster@vmproanc.alcantellab.com, vmproanc.alcantellab.com running ..., Distributed, Remote, Not Registered, TX(14/5/2012 15:31:57)RX(31/12/1900 16:00:00) Q(0)Unitiialized, Alive" - this repeats several times then...
2) "SyncTraceMsg slave vmprofai.alcantellab.com sending "Create: Master" to vmsyncmaster@vmproanc.alcantellab.com" then error appears

This message disappeared3) bool CSMTPConnection:Connect(SQLU ... ): Could not connect to the SMTP server on port 25 ... name or service unknown.

System A Webcontorl Erros
1) can't find correlating errors.

 

[TheSmash]

Yeah, I couldn't get it working on 8.0 either. Would have raised a bug but then 8.1 BETA came out and problems went away! ;-)

Everything above looks correct, I think it's just broken in V8.

ACSS (SME)

One of these days everything will work as it should, and then we'll all be out of a job!

 

[amriddle01]

That's good, higher version fixes = free upgrades

 

[TheSmash]

Andy - lets see if it still works by the time it goes GA!

ACSS (SME)

One of these days everything will work as it should, and then we'll all be out of a job!

 

[bmast]

Ok guys, thanks for the feed back. I feel both validated and disappointed, I expect better of Avaya if this is truly inoperable in release 8. I am repeating the testing on release 7. How do I go about getting the release 8.1 BETA, or better yet do I even want it. This lab testing is for a 5 location / 300 phone SCN deployment with distributed servers. Don't want to downgrade to 7, nervous about deploying with a BETA, but I need a path forward. Suggestions?

 

[tlpeter]

Put 7.0 in if you really need it.
You won't get in to the trials for just this option.


BAZINGA!

I'm not insane, my mother had me tested!

 

[bmast]

WOW! It is broken in release 8. Repeated the tests above on release 7 and Distributed VM worked flawlessly.... what a shame. I am a SE so I have used the line, "well because it is Avaya, they have much more rigorous testing standards. They have the resources to make sure their software is leading edge not 'bleeding edge'." Guess I have to eat my words...

 

[bmast]

I have a fully programmed SCN system, installed on site with 4 servers. We had the first cut scheduled for Friday, this issue is the only untested feature stumbling block. What are the ramifications of running my 5 IP office systems on 8.0.42 and running release 7 on the 4 voicemail servers? It seems to work fine in the lab, is this a supported configuration or is it likely to bite me in the bum?

 

[TheSmash]

Avaya Line - "It's not supported"

My line - "It will probably work but you may encounter a few issues here and there. You won't get any support for it so you'll have to work out the bugs yourself."

If it works do let me know!

Good luck!

PS - I've been testing this since 6.0 beta and it didn't work in that. It's been a bit sporadic and I've only deployed 3 in the field (2 x v7.0 and 1 x v6.1). Came across the issue in v8 when I deployed it on our in-house system. It just didn't work. A week later I had 8.1 and that seems to have sorted things.

ACSS (SME)

One of these days everything will work as it should, and then we'll all be out of a job!

 

[bmast]

Thanks for the help. It is definitely hit or miss. The project must move forward so I will let you know if there are unexpected feature interactions that I didn't discover in testing from mixing release 7 voice mail with release 8 systems.