Tivoli Distributed Monitoring Monitors

• Problem Monitoring the Status of a Perl Script on AIX 4.3.3 [25/7/02]
• wlseng Command Returns Value "down" when Using the Universal Application Status Monitor [24/7/02]
• wlseng Usage Confusion [25/7/02]

Problem Monitoring the Status of a Perl Script on AIX 4.3.3

We are trying to monitor the status of some Perl scripts on an AIX 4.3.3 box. The scripts should always be up and running. If I do a "ps -ef | > grep scriptname" I get a return. So I thought I could monitor this as a daemon or process using DM. But it turns out I can't.

Is there any way I can monitor whether or not these scripts are running through DM? Or do I need to write a script which will do a grep and then forward an event to Tivoli via (w)postemsg if it is not?

Contributed by: Laura Confal
So I assume your process name is something like "perl myscript.pl".

Within the Perl script, you could alter $0 (the calling script name) to be equal to your scriptname. Something like $0="myscript.pl";

I recall having problems resetting $0 on some Unixes (HPUX 10??) but it usually worked out just fine.

Otherwise, a custom script would be in order.

In this particular case changing $0 within the script to $0="myscript.pl" did the trick. Now DM can monitor the script as if were a process running.
 

wlseng Command Returns Value "down" when Using the Universal Application Status Monitor

I am testing out the DM 3.7 Universal Application status (appStatus) monitor. For the test I start Notepad on the Endpoint and add monitors for both Notepad and Notepad.exe.

However, the wlseng command returns value "down" for both of the above.

Contributed by: Ken Matheson
Yes, this is unfortunately correct. You need patch 3.7-DMN-0004, which in turn requires patch 3.7-DMN-0001.  This fixed the problem for me on a Windows 2000 Advanced Server Endpoint.

The monitor is running ntprocinfo -c -s "program". Try it from the command line, and you'll see that you don't need the .exe, just Notepad.
 

wlseng Usage Confusion

I'm a bit confused about the documentation on the wlseng command. The syntax is:

wlseng [-l] [-d] [-z] [collection] [monitor] [instance] [node]

I have an "NT_Memory AvailBytes" monitor running on an Endpoint. "wlseng -l MyEndpoint" gives me everything, as expected.  However, from the documentation I would expect to be able to narrow my search doing i.e. "wlseng -l NT_Memory AvailBytes MyEndpoint"

But All I get is nothing:

Timer:

Monitoring probes:

Responses:

Am I doing something wrong ?

Environment:
FWK 3.7.1 with patches 1,2,3,7,8
DM 3.7 with patches 1,4

Contributed by: Scott Fleming
This command works very similar to the wrunprb command in that you need to provide wildcards for the parts you don't care about. eg:

wlseng -lz "NT_Memory" "AvailBytes" "*" MyEndPoint

This command will provide a listing of all AvailBytes probes from any profile on the Endpoint.