We have a scenario in which the debmas idocs are sent in batches of 5000 idocs in one shot to XI. XI updates 3 legacy systems and the whole process causes high CPU Load.
I decided to use bpm to collect idocs and then process them one by one or may be in chunks of 20 or so.
After doing bpm collect by time I noticed that XI is starting a new bpm process for every incoming idoc. How do I make it to collect idocs? I have gone through the weblog/help explaining the collection of idocs based on Time in the system patterns.
Any ideas what is wrong? I have a list in which I append the individual idocs one by one.