Files on original subvolume are corrupted when writing into snapshots

[bhzhu203]

vesion 5729a4cb0775a8bc15de33fbf481a719dc01a5ee

The original subvolume keep runing a mariadb instance whith slave writing. At the same time a script make FS snaphots every 10mins . Another script delete snaphots and keep snapshots in 2 days. As now on , it works well.

But when I running a recovery instance on a snapshot ,after 20 minutes , the original mariadb has files corrpution. docker run -it --name 1684381801 -p 3306 -v /data/snapshots/1684381801/mysql:/mysql mariadb-copy /start.sh

[koverstreet]

Does the recovery process write to the snapshot?

[bhzhu203]

Does the recovery process write to the snapshot?

Yes , a new mariadb process

[koverstreet]

Fixed by 3012736674