Hi.
We have a situation where NAS 'hangs' periodically. MS has suggested a fix, but it will be a while before we can implement it.
In the meantime, it would be helpfull to know when NAS hangs. By 'hang', I mean that it simply stops processing. When you login to the server an error comes up that say's there was a problem - but you have to login to get the message.
Does anyone know of a simple way to detect if the service is running? (And, by running, I mean actually running?)
One approach may come from this: NAS (is supposed to) poll for files to process every ten minutes. And, every polling interval it generates events that can be seen in the Event Viewer. Maybe it is possible to write something to poll the event log and raise an alarm if no NAS events are seen in the last ten minutes or so. (Idon't need a complete soluton, just a 'nudge in the right direction would be great.)
Thanks in advance!
0
Comments
Independent Consultant/Developer
blog: https://dynamicsuser.net/nav/b/ara3n
2) Which DB? SQL or Native?
We experienced some "hanging" with NAS on 64bit OS communicating with MS SQL 2005 64bit on same machine. After we disabled the Shared memory protocol on the MS SQL and forced SQL to use only TCP/IP, the problems was gone...
MVP - Dynamics NAV
My BLOG
NAVERTICA a.s.
It's running on Win server 2003 and it's the NAV database server.
The issue is NAS hanging while executing an XMLPort, and MS has provided an update to 5.1 which they say has solved similar problems for other users.
RIS Plus, LLC
MVP - Business Apps
I guess I could run a 2nd instance of NAS and have it monitor the sessions table - but that's getting kinda silly...
Independent Consultant/Developer
blog: https://dynamicsuser.net/nav/b/ara3n
If you doesn't use the timer, for example because u use Port communication you can check with a little "Hello NAS" routine (for example every 30min) if your NAS work or is hanging.
There are some ways to solve this. One other is the solution with the session view, but here you doesn't know if the NAS run or not (because he hangs for example in a confirm ](*,) ) u know only that he has a session.
In one of my customer solutions i created 2 NAS. These two NAS talk. If NAS two, in status "IAMTheGuardianNow", check that NAS one (in Status "IAMTheWorkerNow") doesn't answer after some questions, he kill the NAS one and change his own status to "IAMTheWorkerNow".
Then Nas one restart and is now in status "IAMTheGuardianNow". So i simulate a NAS cluster. If NAS two kill NAS one (or NAS one kills NAS two) he send a sms and mail to the admin. If the Kill -> Start -> Kill -> Start situation comes ever again, for example 3x successively (because both NAS run into the same error) , bot NAS stop his work or cancel the job and send also a mail / sms that a specific job create a error.
Regards