1 of 1 people found this helpful
Taking a step back, using conditions in this scenario would not be my first choice. If you don't care about the sequence in which the jobs run, only that they don't run in parallel, then using a Control Resource (set to exclusive) is much easier. Just define the same Control Resource (not forgetting the exclusive setting) in each job, remove the conditions and you'll never have to worry about the jobs running at the same time again.
If have lots of jobs like this (much more than 4) then you might want to define a Workload Policy (makes things more efficient for the Control-M Server) but that is something to consider in the future. There are also Quantitative Resources, but again, let's keep things simple for now.
You can then use the "alert on second failure only" method as per the document that you linked to.
2 of 2 people found this helpful
Hi Steve, I came across a similar thing at a previous company where file transfers were failing sometimes on the first connection but usually successful on the 2nd connection. What we did was, on failure, force in a non cyclic file transfer job to do the transfer. If this is successful then it runs once and the original cyclic job carries on at the next scheduled run. If it fails then it raises a ticket. As Mark suggested, using a Control Resource would be the better than conditions
I'll look into a Control Resource. We don't really care which order the jobs run in.
Initially, this particular flow was 4 individual AFT jobs that ran every 15 minutes at the same time. We changed it to one singular flow to mitigate the simultaneous connection issue with the vendor.
Thank you, Steve!
I'm going to look into Mark's suggestion further, but I can see this as a viable option, as well.
Circling back around to this.
I've been looking at control resources and also workload policies combined with quantitative resources.
I'm assuming if I go the workload policy route with quantitative resources, I wouldn't have to configure a control resource.
Looking at both of those options, I still don't see how I can use rerun on first failure while keeping the the jobs running every 15 minutes, 24x7.
Did I miss something?