Automated Report from OnDemand Client

[Mark@M]

I'm fairly new to using the CMOD suite of tools and was wondering if the following is available ?
Occasionally, I go into the OnDemand Client and go to the System Log Folder.  Once in the folder , I'll do a search on MSG_NUM = 13 , to see if there were any DB issues from the last 24 hours.

Question - Is there a way to run this query thru an automated fashion, either thru a windows environment or mainframe job to produce the same report and have that archive as a daily CMOD report ?

The current way our team handles this seems a little clunky.   We have a series of sort steps running on our mainframe to first determine the system log table name, as this changes over time.  Once determined, we'll query the system log table , extracting the timestamp and and a partial piece of the MSG_NUM(1st 100 bytes). 

Using the reporting tool in an automated fashion seems like a much easier option , if available. 

[Justin Derrick]

Hi Mark.

There aren't many automated reporting tools for CMOD.  IBM does have one called ECM System Monitor, but it covers the entire ECM portfolio, so may be expensive if all you want is some summary info from CMOD.  Some IBM Business Partners offer solutions like connectors to Enterprise Monitoring or standalone reporting products.  I have a tool called RAPTOR4 (the website it currently being rebuilt) just eMail me if you want details. 

-JD.

[Greg Ira]

You don't mention your environment beyond mainframe.  What environment are you on? We're on z/OS and we have System Automation trap crucial CMOD messages like ARS0013E and email them to us when they occur though we could log them in a dataset and send them once a day if need be.
 Also instead of trying to determine the current table name use an ARSDOC query.  We use something like this which you could modify to your requirements.  This lets CMOD determine the correct table for you.
/usr/lpp/ars/V9R5M0/bin/arsdoc query -h OND05P -G "System Log" -f "System Log" -u xxxxxxxx -p /SYSTEM/etc/xxx/arsxxxx.stash
-i "where time_stamp_dt > TIMESTAMP(CURRENT DATE - 7 DAYS) and msg_num= 201"

[Justin Derrick]

Greg, you just reminded me about arslog.

Mark, you can go look up the Logging User Exit, called arslog.  You can write a small shell script on the server that gets executed each time CMOD would normally write to the System Log.

Inside that script, you can compare message numbers, and you can perform an action (send an eMail, send an alert to a pager, etc.) notifying you of whatever issues.

The only caveat is that because this script will run each time a System Log message is generated, it can create a lot of load on your server if it gets particularly busy.  I've seen logging exits bring down servers because there were so many messages being written each minute that the scripts became backlogged, and the CMOD server became unresponsive.

Check it out, and let us know if you have any specific questions.

-JD.

[Mark@M]

Thanks Greg and Justin !!   Will take a look.