Our company recently underwent a merger and we took over a company, this means our model now has become bigger overall - New SKUs, New Customers , new Production lines etc.
The size of the model:
before go-live with merged entity: ~50GB
After go-live with merged entity: 79 GB
Agree and understand that the model size increase is because of the increase in general data volume.
In this context, we've MRP Optimisation Process (Material Requirement Planning) in our supply model running twice a day, triggered by integrations.
This is a pretty lengthy process containing 120 actions in sequence, most of which are "import" actions to cells of the same module (i.e. source and target module is same)
MRP Optimisation process runtime pre merger data: ~20-22 Mins
MRP Optimisation process runtime after merger :~40-45 Mins
The run time has nearly doubled.
I'm trying to identify the root cause for it.
Some more detail on the MRP process, around 90% of the actions of this process are import actions within the same module (same source and target module ).
We have a manual exclude flag for items to be excluded from the MRP optimisation process, this exclude is toggled true for all of the newly added products, essentially any new products/items , which was added as part of the merger is excluded from the MRP process run, and these items are not included in the import actions saved views.
Given this, my expectation/understanding is that the runtime of import actions wouldn't change much since the number of records in the saved view hasn't changed much before and after merger, meaning there is still only the usual number of records to be processed as part of the import action, yet we could see the runtimes of each actions has nearly doubled.
What could possibly be the reason for this.
A thing to note here, the module involved in most of the import actions is dimensioned by a list "Final Product List" and this Final product list is the biggest list of the model and so is the module dimensioned by this list.
The module involved in the import actions of MRP is the biggest module of the supply model(the imports have the biggest module of the model as its souce and target), However I think this shouldn't really matter given the number of records in the import saved views are around the same range as before the merger.
This final product list has new items added as part of the merger, but these added items are what's excluded from the import saved views of the process. so the list has got bigger but not the saved view, will the system still traverse the whole list instead of only the saved view records. Can this be a contributing factor to the increased run times?
What else could influence the runtime of the import actions other than the saved views.
I noticed that few line items of this module have high calculation effort when seen in the saved views, can optimizing these line items to have low calc effort have an improvement in the process run times. I think no?, from reading i understand calc effort is the time taken for a line item to calculate relative to other line items when the model is open. Can optimizing this parameter improve process runtimes.
Any suggestions/help in understanding the root cause of increased process runtime is appreciated!
