RE: issues with ITM for BI (Websphere MQ)

1. You can't distribute those resource models to TMAs (Endpoints) - you have to distribute them to the appropriate new kind of object that is created during the discovery task.

 

2. Can you run the 'wdmlseng' command with at least the '-verbose' option to show us what the actual output is. Other useful flags are:

 

-r resource_model_name -i indication_name

 

You just need to run some more commands and provide the exact output (changing any names appropriately for privacy issues).

 

3. I'm not sure.

 

-- Frank

 

Frank Tate
Frank_Tate@xxxxxxxxxxxx
Gulf Breeze Software
www.gulfsoft.com
phone: 512-426-6189
fax:   208-730-1811
OpenESM Project
www.sourceforge.net/projects/gulfsoft

 

 

---

From: owner-tme10@xxxxxxxxxxxxxxxx [mailto:owner-tme10@xxxxxxxxxxxxxxxx] On Behalf Of Phillip Rhodes
Sent: Friday, February 25, 2005 4:53 PM
To: tme10@xxxxxxxxxxxxxxxx
Subject: [tme10] issues with ITM for BI (Websphere MQ)

Hi all, I have a couple of questions 
I'm hoping someone 
can shed some light 
on for me. 

I'm running TMF 4.1.1 
with ITM 5.1.2, ITM for BI 5.1.1 and TEC 3.9, on a Win2K 
server.  I have an endpoint running WebSphere MQ on 
Windows XP.  I've run the "Discover Queue Managers" 
task on my TMR server and it did successfully locate the 
Queue Managers in question and I can do things 
like list queues, start and stop queue managers, etc., from the Tivoli 
Desktop. 

However... 

1. If I distribute a monitoring profile 
with the Queue and/or Channel resource models 
marked active, neither the Queue nor Channel RMs will start.  I get 
back a "retrying (110)" 
message from 
both. 

2. The QueueManager resource 
model starts (apparently) but I'm not convinced it ever does 
anything.  In particular, there is an Indication for 
"Queue Manager Unavailable" in that RM.  But 
if I stop Websphere MQ ( by any means, including manually 
killing the process), the indication never 
seems to fire.  I never find an event in TEC based on that 
indication.  I've run wtdumprl to make sure 
it isn't the case that it got sent but simply failed to 
parse due to a problem with the BAROC files, etc. 
No dice.. it just isn't there.  Also, if I load Web 
Health Console and look at the status of the resource 
model's, the QueueManager resource model always shows 
100% health, no matter what. 

3. 
Another bit of quirky behavior: if I stop MQ on the endpoint, the QueueManager 
icons in Tivoli Desktop 
never switch to 
the "stopped' icons. They always show up as "running."  Maybe that's the 
expected behavior 
but it seems odd to 
me.   


Anybody have any 
insight into any of these issues? 


Thanks, 

Phillip 
Rhodes
------------------------------------------------------------------
Pervasive 
Computing - RFID Server Development
E-Mail : prhodes@xxxxxxxxxx
Phone : 
(919) 486-1499
T/L       : 
526-1499
-------------------------------------------------------------------