I have a planning application (8.4.1) which has 1200 eList items. The GTP completes, but the reconcile job cancels repeatedly. When it cancels, I hit GTP again to pick up the reconciliation where it left off. It eventually completes, but not after cancelling 5 times. Could this be an issue with the job server not having enough memory?
What does the planningerrorlog.csv show under the service account's profile?
How much virtual memory do you have configured on each job server? Also, how many processes are you allowing on each job server? Both of these get at whether or not enough resources are being allocated to each individual job to allow it to completed before starting another job.
How many job server threads and what sort of error messages, bulk load failure? dead lock resources?
Stuart
Thank you for the replies, I've determined it is due to the fact that the client does not have a dedicated job server. There is one server which hosts Planning as well as Cognos 8 reporting. The processor gets to 100%. They need an additional server.
I don't know if this will help or not, but a Planning Job server is designed to use 100% CPU utilization.
Ex:
4 CPU job server = ~25% utilization for each processor
8 CPU job server = ~ 12.5% utilization for each processor
etc..
I can definitely agree with you that they should have a separate job server. I just wanted to make sure you were aware of how a job server has been designed.