View previous topic :: View next topic
|
Author |
Message |
tjk3030
New User
Joined: 18 Mar 2006 Posts: 22
|
|
|
|
We have numerous batch jobs sending files to other NDM servers outside of our company. In the NDM process, we have routines to check file status etc. Which is working fine. However lately we've had a problem where the clients server is down & we are not connecting. Is there any way to check the status of the snode server? Some kind of batch thing I can run. How I'm finding out now, is the client calls and said they did not get their files. |
|
Back to top |
|
|
priyesh.agrawal
Senior Member
Joined: 28 Mar 2005 Posts: 1448 Location: Chicago, IL
|
|
|
|
This is a normal practice to check for the return code from NDM step and abend the job in case its not equal to zeroes. It'll solve your problem for cases other than server unavailability as well.
So first check the job logs from a successful and an un-successful job execution to find out if you are getting different return codes (from the NDM step) or not. If the return codes are different then you can make the job fail. And then find out what went wrong. |
|
Back to top |
|
|
tjk3030
New User
Joined: 18 Mar 2006 Posts: 22
|
|
|
|
The failure is occuring at the connect step, so any of the process statements following do not get executed, I have not found anything to check a return code on the "Connect" step. In all of our processes I check return codes & care out the appropriate response. But since the connect fails never gets that fair. I have to go look at the log to see the connect failure. |
|
Back to top |
|
|
priyesh.agrawal
Senior Member
Joined: 28 Mar 2005 Posts: 1448 Location: Chicago, IL
|
|
|
|
hmmm... what about routing log to a dataset and check for specific key words (such as "failed", "Connection failed" etc) and make the job fail if find anything in the dataset. |
|
Back to top |
|
|
tjk3030
New User
Joined: 18 Mar 2006 Posts: 22
|
|
|
|
Well what I've decided to do, is create a job that will submit an NDM job for all the different nodes every few hours, that job will run an NDM to task to create a file on our mainframe. i'll have another job check for the existence of those files, if not found, means the connect step failed, and the create file steps never executed. Should work in theory although kind of mickey-mouse. Actually the problem was very rare happened 2 times in a week, different reasons why NDM could not connect, but net result was the same. A days worth of files was not transmitted & we didnt know till they called us asking where are their files??? Cannot remember the last time it happened if ever did, in the past 5 years. |
|
Back to top |
|
|
|