Monitors RU consumption to maximize short-term usage of MAX RU. #7908
Labels
affects-7.1
affects-7.5
report/customer
Customers have encountered this bug.
severity/minor
type/bug
The issue is confirmed as a bug.
type/development
The issue belongs to a development tasks
Problem
I have a group:
When I only run workload A:
RU Avg is OK because the workload is stable.
But When I manually run the big query SQL like:
The monitoring was not very accurate(on-avg), which led me to mistakenly believe that I was far away from triggering RC control.
But from the slow query:
it cost 8000+ RU, and which let some query need wait in the RC queue
Development Task
enhance the observability.
The text was updated successfully, but these errors were encountered: