I don't know how to explain this well, but I will try.
- I use Google Cloud SQL second edition with 20 GB disk size.
- I have several wp. databases with 166.5 MB Total size.
And right now my Storage usage is 9.52 GB! (With 166.5 MB SQL data...) and increasing still going faster...
What to do?!
UPDATE :
I solve this with :
- I made an export in bucket
- I created a new instance Cloud SQL
- Import from bucket
- And delete instance with problem.
(And changed ip from my applications)
I don't know sure where problem come from but could be a "Storage overhead from binary logs".
Next time will check binary logs with : mysql> SHOW BINARY LOGS;
What I think Google is missing is a purge binary logs! (an easy way!)
UPDATE FINAL :
With binary logs active, storage of your cloud SQL will expand continuously.
For anyone in the same situation, you can edit the instance and uncheck binary logs, after that the current binary logs will purge.
Sorry for my noob problem! :D (I'm a beginner in Server administration.)
Thanks Vadim!
2条答案
按热度按时间mbskvtky1#
If you have binary logs enabled, mysql will make a record of all changes, which is required for replication or point-in-time recovery.
If you have no need for these features, you can disable binary logs which will purge any existing logs from your instance.
If binary logs are enabled, they will not grow indefinitely. Binary logs older than the oldest automatic backup (7 days) are purged automatically.
jfgube3f2#
I had the same issue with our GCP SQL instance. In my case the storage went from 10GB that we first configured to 728GB (with an increase in billing accordingly).
In my case the culprit was archived WAL Logs:
Metrics explorer
Just viewing the current log size did not show this because the logs had already been cleared.
The issue is that GCP has a hidden option:
Automatically increase storage option for GCP SQL
This is checked by default (IIRC). So the storage will increase automatically in events of high log output (in my case this happened when upgrading Postgresql server version.
The catch is that once increased, the storage cannot be decreased, ending up with increased monthly operational costs.