index checkpoint size

56 views

Skip to first unread message

Amel

unread,

Jan 11, 2018, 5:27:51 AM1/11/18

to mongodb-user

I loaded my data to a new database in a Pre-Production environnement and I find that the index size has highly increased compared to the same thing in a développement environnement.

When I comparend the size of each index between the developpement and the Pre production environnement, i find that the difference is on the "checkpoint size".

Is there a method to clean/remove the space taken by the checkpoint during the data loading and not used after.

Thank you for your help.

Kevin Adistambha

unread,

Jan 22, 2018, 10:32:44 PM1/22/18

to mongodb-user

When I comparend the size of each index between the developpement and the Pre production environnement, i find that the difference is on the “checkpoint size”.

A checkpoint is a consistent snapshot of your data, written to disk by WiredTiger. Checkpoint size would be related to the amount of data written to the database since the last checkpoint occured. By default, MongoDB performs a checkpoint every 60 seconds, so the checkpoint size would be related to how much data was written to the database in the last 60 seconds.

In your case, are you seeing this large checkpoint size during the data import process? This would be reasonable since you are writing a lot of data into the database.

Is there a method to clean/remove the space taken by the checkpoint during the data loading and not used after.

Checkpointing is the mechanism used by WiredTiger to persist your data to disk. Thus, there is no reason to “clean” anything.

Please see Snapshots and Checkpoints for more information regarding this process.