Atlas online archive partition based on timestamp

Hello there!
I’m plan to migrate a large amount of data to mongodb with ability to move some data to Atlas Online Archive.
Average size of the record is 1.6Kb.
I saw a video with mongo db engineer explaining how to use AOA by using archiving rule based on timestamp A Day at the Data Lake: Atlas Online Archive with Bob Liles Atlas Engineer - YouTube
So the question is: when archiving rule is based on timestamp does data will be partitioned by this key in addition to 2 keys available for partitioning (resulting in 3 partition keys in total)?
Will be queries effective in cost and speed as if I use for partitioning date-hour key (e.g. 2020-10-24T23) ?

Hi Dmitry,

Yes that’s correct, if you are archiving using a date-based rule, you can choose up to 3 partitioning keys in total, including the date field you selected. For a custom archival rule, you can choose up to 2 partitioning keys . The queries will be effective if you are most frequently querying the archive using the same key. Otherwise, you might push the date field to the lowest in the order of partitioning keys.

Please refer to this blog to know the list of do’s and don’ts for partitioning fields:


1 Like

This topic was automatically closed 5 days after the last reply. New replies are no longer allowed.