PMM Incrementa Backup Size is not correct

Hi,

We are taking daily incremental backup.

Base Image side is 2.2TB. Daily incremental images are creating more than 500GB.

This is not correct. Our daily data growth is less than 10 GB.

Why incremental image size is too huge. Please check and suggest.

Snapshots:

2025-10-01T15:00:05Z 553.95GB [restore_to_time: 2025-10-01T15:00:09Z]
2025-09-30T15:00:04Z 553.46GB [restore_to_time: 2025-09-30T15:00:08Z]
2025-09-29T15:00:05Z 556.38GB [restore_to_time: 2025-09-29T15:00:08Z]
2025-09-28T15:00:04Z 2.15TB <incremental, base> [restore_to_time: 2025-09-28T15:00:08Z]
2025-09-27T15:00:05Z 550.08GB [restore_to_time: 2025-09-27T15:00:14Z]
2025-09-26T15:00:04Z 555.41GB [restore_to_time: 2025-09-26T15:00:08Z]
2025-09-25T15:00:05Z 555.45GB [restore_to_time: 2025-09-25T15:00:20Z]
2025-09-24T15:00:05Z 555.67GB [restore_to_time: 2025-09-24T15:00:09Z]
2025-09-23T15:00:04Z 552.64GB [restore_to_time: 2025-09-23T15:00:08Z]
2025-09-22T15:00:04Z 555.63GB [restore_to_time: 2025-09-22T15:00:08Z]
2025-09-21T15:00:05Z 2.15TB <incremental, base> [restore_to_time: 2025-09-21T15:00:09Z]
2025-09-20T15:00:05Z 553.48GB [restore_to_time: 2025-09-20T15:00:09Z]
2025-09-19T15:00:05Z 561.25GB [restore_to_time: 2025-09-19T15:00:09Z]
2025-09-18T15:00:05Z 557.01GB [restore_to_time: 2025-09-18T15:00:09Z]
2025-09-17T15:00:04Z 557.00GB [restore_to_time: 2025-09-17T15:00:08Z]
2025-09-16T15:00:04Z 557.37GB [restore_to_time: 2025-09-16T15:00:08Z]
2025-09-15T15:00:04Z 559.69GB [restore_to_time: 2025-09-15T15:00:11Z]
2025-09-14T15:00:04Z 2.14TB <incremental, base> [restore_to_time: 2025-09-14T15:00:21Z]

Thanks,

Sumit

Hi @sumitkumar.gupta,
Data growth may be only 10GB, but what about data change? I can have a 500GB database, then INSERT 10GB of data, then UPDATE 200GB of data. Total on-disk is 510GB, but I’ve changed 210GB thus my incremental backup would be 210GB in size, not 10GB.

Hi @matthewb

Thanks for reply. I don’t think, we are updaing 25% of data everyday. Is there anyway, we can check how much data is changed during specific time.

Thanks,

Sumit

You can check this by looking at InnoDB’s redo log sequence number (ie: LSN). Grab the number now, then wait an hour, or two. Look at the number again. Subtract the difference. This number represents the amount of bytes that have been changed in that time period.

show global status like 'Innodb_redo_log_current_lsn';