SAP Noted 1840729: Implemented, program: ZSCHEDULE_FICOINT_PLAN_ADVANCE

[former_member195427]

Dear Experts,

SAP has given us program ZSCHEDULE_FICOINT_PLAN_ADVANCE (as per Note 1840729) and this program contains almost all parameters of program /RPM/FICO_INT_PLANNING.

I am facing some issues while creating variants to schedule a background job to run this program.

Actually, 1) I am not sure what this field' Number of free work processes' is all about? Which value should I enter here to make my variant more meaningful and efficient in running.

2) Item count(job size): What should I keep in this field vacant and run my variant. Is this field has any co-relation with field 'Number of free work processes'?

We are using 3 portfolio's and I have created 3 variants, 1 for each portfolio. Each portfolio contains some 1000 items inside it.

So, kindly help me in setting up these variants for a smooth and efficient job run for this program.

Best Regards

Saurabh

[Former Member · Accepted solution]

Hello Saurabh,

I may not be able to answer you question fully, but I do have the following contribution to make:

We are also using this program which is a "wrapper program" designed to split the Fin integration jobs in a better way. The traditional /rpm/fico_int_planning report has the restriction that you have to maintain variants and quite often have to keep updating these variants as your portfolio grows - e.g. you have a variant for Bucket A but this bucket size keeps growing and now you have to split by type, etc. This was not ideal and hence this new report...

The 'Item count' field now helps to group records into batches - e.g. of 100. This means that if you have 1000 Items and you execute this job, it will create 10 background jobs of 100 each (in sm37 you will see this). We have seen however that all records in table /rpm_d is being considered - even templates, snapshots and other non-relevant Items for this purpose. The code should actually only consider Item class ' ' or '2' that are non snapshot relevant. We enhanced the standard report to exclude these (perhaps SAP has in the meantime excluded these?). Without the enhancement we did (and if SAP hasn't enhanced their select) you will get all records in Item D

Not sure about the number of free work processes (this seems to be "new"), but I can venture a guess that you may set here that you will not allocate all of the existing processes, and that a certain number of processes must be left for other processes in the system. E.g. your system has 20 processes and in this setting you specify 15 meaning that your job will only make use of 5 processes.

Hope this helps a bit

Regards

C

[former_member195427 · Accepted solution]

Thank you very much Chatsworth!

I did utilized the 'job count' field in the similar way as you explained and yes it looks good now except for an instance where I saw 1 or 2  process of this job run are cancelled. It actually happens every time I schedule and run it and I find 98 successful and and 2 cancelled ones.

The cancelled ones shows termination due to MEMORY_NO_MORE_PAGING. So, should I discuss with Basis guy on it if to adjust some parameter or should I increase the job count to 200?

Secondly, While checking SM37 job log I could see 12 jobs created at one instance for 3 variants that I created with item count 100. How these 12 jobs for 3 variants each with item count 100 were run by system?

Also, I observed that once these first 12 jobs run are finished then system waits for few seconds (probably 40/45) and then initiates other 12 jobs and so on..  Is this related to 'number of free work processes'? How can I minimize this gap to 5 seconds only?

Best Regards

Saurabh

[Former Member · Accepted solution]

Hello Saurabh

As far as optimisation goes (number of work processes etc.) it is best to discuss with your Basis team. I can tell you that this report, as a general comment, runs very inefficiently - firstly since it includes all objects in Item D table and, as with lots of other standard PPM code, triggers lots of other processes/calls/methods that may not all be applicable. Hence you may find that the 40/45 second gap that you experiencing might be due to some other stuff finishing off. If you do a SM66 trace you will see lots of things happening, and also depending on your configuration this process may take longer - e.g. do you have financial views configured as integrated and also rolls up to Buckets.

I am not quite getting what you mean with the 12 jobs and 3 variants paragraph?

Regards
C

[former_member195427 · Accepted solution]

Hello Chatsworth,

With 12 jobs and 3 variants I mean that I created 3 different variants of this program and assigned these variants as 3 steps (each with item count 100) in scheduling a background job with 3 steps.

When I saw SM37 Log I found a main job run with 12 child job runs (all with /RPM/FICO_INT_PLANNING). Also, once these 12 jobs are finished then another 12 job runs after that ad so on..until all the items are processed.

So, could you please suggest me the reason why it happened? Actually I was expecting only 3 job run for 3 variants that I created.

Best Regards

Saurabh

[Former Member · Accepted solution]

Hi Saurabh

When you created the variants - did you specify details such as Bucket ID? Did you - for example, create variant 1 for Bucket A with job size 100, variant 2 for Bucket B with job size 100, and variant 3 for Bucket C with job size 100?

If you did something like the above, then job 1 will select all Items (versions, templates, items, etc) for Bucket A. If Bucket A therefore has 300 records then job 1 itself will run 3 batch jobs. Same logic applied to Bucket B and Bucket C and you will therefore however many jobs that are relevant.

Not sure this makes sense or speak to your example?

Regards

C

[Former Member · Accepted solution]

Hello experts,

I would like to know how to run the report ZSCHEDULE_FICOINT_PLAN_ADVANCE for a single item.

While executing the above report for a single item id, we found that the report was running for the entire bucket instead of running for the single item.

Please clarify on,

1.What selection screen parameters are to be used in ZSCHEDULE_FICOINT_PLAN_ADVANCE.

2.How to run the report ZSCHEDULE_FICOINT_PLAN_ADVANCE for a single item or a subset of items/projects.

Thanks in advance.

Regards,

Aishwarya.

[Former Member · Accepted solution]

Hi Geeks,

Could anyone please provide a solution to my above issue quickly ?

Thanks & Regards,

Aishwarya.

[former_member195427]

Hi experts,

Also to add that this program is working successfully in ECT. But in Production program: /RPM/FICO_INT_PLANNING. was failing with memory allocation error that's why we went ahead with SAP note and its program.

Best Regards

Saurabh

[former_member195427]

Any pointers on it Experts?

Best Regards

Saurabh