An instance of our Google Cloud SQL instance filled up to 10 TB of disk space (thank you auto expansion ;) with a lot of
[Note] InnoDB: page_cleaner: 1000ms intended loop took 4546ms. The settings might not be optimal. (flushed=9360 and evicted=13008, during the time.)
log entries. We restarted it and now disk space is back to ~250GB. However the instance still seems to have 10TB of SSD attached and I could not find a way to reduce it back to 500 GB in the ui (seems only storage increase is supported). Any pointers to reduce disk usage?