We have about 200 long running microflows in our production environment that won’t end even after attempting to Kill the request. The longest running microflow is a scheduled event that occurs daily that was attempting to delete a few hundred objects (along with the associated objects with cascading delete), and it’s currently at 70 hours; it’s hung up on a delete action. Is there any way to kill this process aside from restarting the app? This issue has occurred in the past and killing the longest running flow allowed the rest of the long running microflows to finish, but this time none of them are able to be stopped. The rest of the long running flows are stuck on various database actions, i.e. committing or deleting objects.
I think you are looking at a restart here.
I wonder if these microflows are manipulating the same data so causing conflicts? It sounds like refactoring your microflows to work in small batches is worth looking at. Try looking at the Queue or ProcessQueue modules on the App Store as this will help break the work down into smaller units that can be completed quicker and in their own transactions.