In the recent storage migrations of pallets (example above) we return the MAXIMUM_BLOCK_WEIGHT for the migrate() method. This creates a weird output when you migrate multiple instances in the runtime, example try-runtime output:

2021-10-18 19:34:55 TryRuntime_on_runtime_upgrade executed without errors. Consumed weight = 5500100000000, total weight = 500000000000 (11.0002)

My understanding is that in this case weight being greater than max-weight doesn't matter as it calls register_extra_weight_unchecked for on_runtime_upgrade (executes regardless of weight) and therefore weight is just a placeholder to ensure nothing else other than the storage migration will be executed that block.

But doesn't that mean that if a storage migration happened to take longer than a block to execute using on_runtime_upgrade() that would stop block production as it's weight is unchecked? I could be confused so any clarification would be helpful :)