High Concurrent User Count - Best Practices?
Posted on 2008-10-15
I have a database that maintains about 3500 - 5000 active records, with basic maintenance (decompile/compact/repair & archive) run daily. It has 30-40 concurrent users. All users have to update records, though none delete or add records. The users do not ever overlap, so they do not access a record that is in use by someone else, but we're at the point where the BE gets corrupted a few times a day now.
There is a main data table in the back end on the server, with front end MDE's distributed to the agents. There is also an import database that runs VBA code to update the data table hourly with new records from an IBM-3278 terminal interface via screen scraping.
What I'm looking for are a list of best practices (other then not exceeding 8 users), tips or tricks that will minimize corruptions.