I am currently working on two Records Management Implementations, the smaller of the two repositories has tens of millions of records, the larger has over 170 million records. I am running into issues using workflow activities to assign retention metadata to the records in the repository (assigning filing date and disposition instructions for 360,000 records took 3.5 days). I attempted to assign to the folder, but it gives a time out error on anything but the lowest level folder (of which there are thousands, maybe tens of thousands). All this hasn't even gotten to the point of searching the repository for records ready to be dispositioned (of which there will usually be tens of thousands at the smaller customer, but over a million each month at the larger customer).

I have done several brainstorming sessions and almost every solution we come up with avoids using the RMM in some way to accomplish the tasks.

Does anyone have experience with implementing a records management program at very large scale that could provide guidance on workflow optimization for assigning metadata, search optimization, destroying millions of records, and any other aspects of an implementation of RMM at such a large scale?