Monitors RU consumption to maximize short-term usage of MAX RU.

[nolouch]

Problem

I have a group:

When I only run workload A:

sysbench --mysql-user=user1 --mysql-host=10.2.12.53 --mysql-port=32335 --table-size=10000000  oltp_read_only run --time=1000 --report-interval=1 --threads=4  --mysql-db=test --tables=32 --rate=2

RU Avg is OK because the workload is stable.

But When I manually run the big query SQL like:

select count(*) from test.sbtest1;

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.

[seiya-annie]

/found customer