Jobs and Schedules .. in MF environment

dan_varma · New User Joined: 09 Jan 2004 Posts: 3

Hi,

I am working on a production support environment in legacy systems(not MF/its mid range system)..It too have similar kinds of softwares like MF.

My question is how to handle production support in Mainframe environment.

Here are the steps I am giving on my current job, Can anybody having experience in M/F production support can give me the equalent in MF.

1.Get the pager on which system/account/job name and job number abended
2.Log on to that system, and we have one utility (lists all the successful/abended/hold) jobs.
3.while giving the job number system displays the standard list
4.find out the error,correct the error..and either re run from the beginning/rerun from the step which has abended/cancel the job it self..etc nature of the job .
5.wait till the successul completion fo the job and close the call in our tracking sytem.

sandip_datta · Posted: Sat Jan 10, 2004 2:44 pm

I am giving production suppoet in MF area for our Client Citibank,Tokyo. I want to add some more along with your points -
1. Before rerun / Skip / or recovery ensure if possible that it will not affect in the batch next. Before rerun always restore files from backup which are partly updated by the job. Before rerun from particular step, make sure that there is no temporary dataset in the JCL.
2. If it is online support, get confirmation from other associated or partner system if any.
3.Ensure that the same error should not happen once again.
4. Above all always try to understand that how your decision is affecting client business.

Regards,
Sandip.

NILESHK2302 · New User Joined: 13 Feb 2004 Posts: 14 Location: hyd

Hi ,

Just to add few things:

1. First of all know the severity of abend. If the abend is so severe as in case of online screens getting affected then inform the users/business in time before starting for a solution.

2. Make sure your SLAs are not missed. As a situation if you have some poduction update jobs and report jobs competing to run and you are near your SLA limits then give preference for update jobs rather than trying to wait for report jobs completing. In the scenario mentioned you can even place a hold on all the report jobs.

3. When you have an abend first of all try to get maximum info you can from the listings,sysout etc. if you dont know what the abend code does mean then in mainframes type qw abend-code. You will get info. When you are confident of restart make sure that you have checked any temporary files getting created before the step u are going to restart. Check if any GDG versions are used , have there beeen any Vsam files being used.

4. Give always an update on the status of abend resolution and keep a log of the abends that come up.

5. To my observation mostly abends occur when there is some change that went in previously , check for any changes went in , system updates done etc. This is very useful when u encounter online abends.

If any more infor can shoot up your questions regarding prod support... 8)