Watchdog Timeout/Forced Reset Merlin Messaging 1

[tdaugirdas]

This one is for any ex-Tier 3/4 Lucent/Avaya engineers.

Have a large (3-cabinet) Magix/Legend system running Magix R4,V7 - with an R4 Merlin Messaging.

Merlin Messaging does spordic "warm" resets during the workweek - interestingly NOT on weekends. Can happen several times a day - skip several days - happen next day once - then several times again. No particular pattern.

Processor error logs are "clean". Merlin Messaging error log - shows a "Watchdog Timeout" error - followed in a few seconds by a "Forced Reset" error - every time the system does a reboot.

Have replaces the messaging twice - same issue. Have put the messaging in a different carrier and slot - same issue.

Cisco routers use the "Watchdog" timer to prevent endless loops in some of their CPU functions - and the timeout does cause a router reset - but what this has to do with Merlin Messaging - I have no idea.

Has anyone ever run into this?? And - what can I try next??

Tom Daugirdas,
President
STCG, Inc.
stcg.com

 

[telephonydave]

Something similar. When a call was rcvd to the voice mail A/A that was blocking CLID on a PRI the A/A would not answer and soon after reset. You might try to place a call with CLID blocking and see it the A/A answers. It sound unrelated to your problem but I have run into this twice with a MPLS PRI.

 

[tdaugirdas]

Hi telephonydave - that sounds interesting. It may be worth a test to see if placing a call to the system using callerID blocking - causes the reset. If so - I wonder how one could stop it from happening.

Tom Daugirdas,
President
STCG, Inc.
stcg.com

 

[telephonydave]

Tom - I used the Monitor PRIGON1 command in WinSPM and monitored the D channel. Each time a call came in with "ANONOMOUS" instead of a number, if it reached the Merlin Messaging it would start a reset. Sometimes it took two to three consecutive calls before it sould start locking up all ports and then a reset would follow. I had no luck working with the provider.

 

[tdaugirdas]

telephonydave - Thanks - I appreciate the info - maybe it will help in this weird case.

Tom Daugirdas,
President
STCG, Inc.
stcg.com

 

[tdaugirdas]

telephonydave - just wanted to THANK YOU again - you were 100% correct - an "anonymous" or "Caller-ID blocked" call hitting the AA caused the voicemaiol port to lock-up - and shortly thereafter caused the messaging to reboot. Checking the error log in messaging - showed the "Watchdog" time-out error and forced reset. Just wanted you to know!!

Tom Daugirdas,
President
STCG, Inc.
stcg.com

 

[merlinman]

Tom - Another tech I have been working with has verified the same failure.

http://home.comcast.net/~merlinman

 

[tdaugirdas]

Jack: Hopefully - having the Telco "block" anonymous calls - will solve the issue. This was the first time I have ever run into this situation - must be one in a few thousand!!

Tom Daugirdas,
President
STCG, Inc.
stcg.com

 

[msi12]

To Tom --

I am having the same issue.
I did a call with *67 6 times to the Merlin Magix 007 MLM and the voicemail module rebooted.

The client setup is an ATT T1 circuit. Setup as a T1 flex circuit with 11 voice channel and internet going into a Cisco 2811. 1 daughter card out to Lan and the other daughter card to 100DCD board.

Did having Telco block anonymous calls work for you?
If not, what solution should I take?

Howard

 

[tdaugirdas]

The only other solution - if they can't block the calls - is to step down to an R1.1 Merlin Messaging - which did not have the problem - ot at least so I have been told.

Tom Daugirdas,
President
STCG, Inc.
stcg.com

 

[msi12]

If you don't mind. Can you tell me where you found this info from?

 

[bryandj23]

Interesting; maybe even a bit ironic...

I've read this thread in the past, but just now realized (now that we have this upon our heads) that msi12 has (what sounds like) the same solution AT&T is proposing us - a flex T1 into a Cisco 28xx.

Still wonder (in my other thread) what programming changes would need to be done.

AND, after reading this again, I wonder if we'll be plagued by this MMesg Watchdog Timeout if we go the IP Flex route.

 

[tdaugirdas]

OK - I didn't realize you had an 007 MLM with the same issue as with the Merlin Messaging R4- this is news to me.

I heard from another tech that downgrading from an R4 Merlin Messaging to R1.1 - solved at least his issue.

A different tech - switched from 5ESS switch-type to DMS-100 - and that solved the reboot issue on anonymous calls - but he had to go back to 5ESS because of too many dropped calls - so I am not sure he ever resolved the problem.

Is the flex circuit configured as 5ESS or DMS-100? You could try switching - to see if the problem gets resolved.

Tom Daugirdas,
President
STCG, Inc.
stcg.com

 

[msi12]

Thanx tdaugirdas for getting back.

I'm not sure what switch type settings Flex-T1 circuit has. I'll call ATT to find out. Currently the Magix is set to DMS-100.
I'll try the switch and see what happens. Will post back.

Thanks

 

[tdaugirdas]

Hopefully they have 5ESS available - to see if that will fix the issue.

Tom Daugirdas,
President
STCG, Inc.
stcg.com

 

[msi12]

I just talked to ATT and there switch type is 5ESS. So I set my to 5ESS. I tested it and still not working.

I'm going to call ATT for assistance.

 

[msi12]

Great news.

We finally got it fixed after making blocked and unblocked calls to vmail system.

Provider found the problem.
You (Tech) need to let the person who is translating the router to include a rule.
Have the router translate (anonymous; unavailable; and restricted numbers) to digits.
Merlin Magix does not have the ability to translate CLID blocking.

Added translation rule to router below so PBX Merlin can interpret digits instead of letters like 'anonymous'



voice translation-rule 1

rule 1 /^\+\(.*\)/ /\1/

!

voice translation-rule 10

rule 1 /^\+1\(.*\)/ /\1/ ç=== added translation rule

rule 2 /^\([0-9]\)\(.*\)/ /\1\2/

rule 3 /.........../ /1234567890/

rule 4 /........../ /1234567890/

rule 5 /........./ /1234567890/


Good Luck

 

[msi12]

This ç=== added translation rule (This did not come out correctly after I posted)

should be a arrow just pointing to the translation.
<=== added translation rule