Cloud SQL Read Replica is growing unexpectedly beyond the source DB sizing.

[Kyle_Herrmann]

A few months ago we migrated our MySQL DB into CloudSQL and have been running a read replica, since day one there. 

Recently, we've run into an issue where the replica grew insanely large. Like, from <1TB to >30TB. It was caught before it became a major issue, and thankfully, being a replica, we were able to blow it away and recreate it easily. I did track it down to being the "tmp_data" data type that exploded. 

Now, I'm seeing a much smaller spike of 1TB in read replica growth, and it's attributed to the "Other" data type. 

Is there any information on what "Other" represents?

[Andre_Fiesco]

Other data type:

This data type is anything using disk storage that is not being captured in the logs as space used, significant growth (anything over a TB) is unusual and if the growth continues to be a problem, then a Google support case should be opened using the following link[1]. 

[1]: https://cloud.google.com/support-hub