On this page
Feature unavailable in Serverless Instances
Serverless instances don't support this feature at this time. To learn more, see Serverless Instance Limitations.
You can configure data in a collection to be archived by specifying an archiving rule. The archiving rule for a:
Time series collection is a combination of a time that is used to determine when to archive data and a numeric value representing the number of days that the Atlas cluster stores the data.
Standard collection can be one of the following:
A combination of a date that is used to determine when to archive data and a numeric value representing the number of days that the Atlas cluster stores the data.
A custom query that is used to select the documents to archive.
To configure your Atlas cluster for online archive:
Create an archiving rule by providing the collection namespace and the criteria for selecting data to archive in the collection.
(Optional) Specify commonly queried fields to partition archived data.
To create an online archive for a cluster using the Atlas CLI, run the following command:
atlas clusters onlineArchives create [options]
To watch for a specific online archive to become available using the Atlas CLI, run the following command:
atlas clusters onlineArchives watch <archiveId> [options]
To configure an online archive from the API, send a
POST request to
the onlineArchives endpoint. If the
cluster already has an
Active online archive with the same
archiving rule for the same database and collection, the operation will
fail. However, if the existing online archive is in
Deleted state, the new online archive is created and its status is
Active. To learn more about the syntax and options, see
To configure an Online Archive, in your Atlas UI:
If it is not already displayed, select the organization that contains your desired project from the Organizations menu in the navigation bar.
If it is not already displayed, select your desired project from the Projects menu in the navigation bar.
If the Database Deployments page is not already displayed, click Database in the sidebar.
Specify the collection namespace, which includes the database name, the dot (
.) separator, and the collection name (that is,
<database>.<collection>), in the Namespace field.
You can't modify the namespace once the online archive is created.
Specify the criteria for selecting documents to archive for the type of collection you want to archive.
Atlas runs an index sufficiency query to determine the efficiency of the archival process. If the number of documents scanned to the number of documents returned is 10 or more, the query result triggers an
Index Sufficiency Warning. This warning indicates that you have insufficient indexes for an efficient archival process. For date-based archives, you must index the date field. For custom criteria that use an expression, Atlas might first convert a value before it evaluates it against the query.
Optional: Specify how many days you want to store data in the online archive and a time window when you want Atlas to run the archiving job.
(Optional) Specify a Deletion Age Limit.
Atlas Online Archive data expiration is available as a Preview feature. The feature and the corresponding documentation may change at any time during the Preview stage.
By default, Atlas doesn't delete archived data. However, if you specify the Deletion Age Limit, you can specify between
9125days (25 years) to keep archived data. Atlas deletes archived data after the number of days you specify here. This data expiration rule takes effect
24hours after you set the Deletion Age Limit.
Once Atlas deletes the data, you can't recover the data.
(Optional) Specify a Schedule Archiving Window.
By default, Atlas periodically runs a query to archive data. However, you can toggle the Schedule Archiving Window to explicitly schedule the time window during which you want Atlas to archive data. You can specify the following:
Frequency. You can choose to run the job every day, on a specific day of the week, or on a specific date every month. If you wish to schedule the data archiving job on the 29th, 30th, or 31st of every month, Atlas doesn't run the archiving job for those months that don't include those dates (for example, February).
Time window. Select the period of time during which you want Atlas to run the data archiving job. You must specify a minimum of two hours. If a running job doesn't complete during the specified time window, Atlas continues to run the job until it completes.
Specify the two most frequently queried fields in your collection to create partitions in your online archive.
Archive must have at least one partition field.
The value of a partition field can be up to a maximum of 700 characters. Documents with values exceeding 700 characters are not archived.
Choose fields that contain only characters supported on AWS. To learn more about the characters to avoid, see Creating object key names. Atlas skips and doesn't archive documents that contain unsupported characters.
Choose fields that do not contain polymorphic data. Atlas determines the data type of a partition field by sampling 10 documents from the collection. Atlas will not archive a document if the specified field value in a document does not match values in other documents in the same collection.
Choose query fields that do not have high cardinality unless you always use those fields in your queries. Query fields, such as
_id, with possibly unique and large number of values can cause operations such as
countto open all partitions resulting in high latency.
Choose fields that you query frequently and order them from the most frequently queried in the first position to the least queried field in the last position. For example, if you frequently query on the date field, then leave the date field in the first position. But if you frequently query on another field, then that field should be in the first position.
For fields of type
string with high cardinality, Atlas
creates a large number of partitions. MongoDB doesn't recommend
string type fields with high cardinality as a query field.
Atlas supports the following partition attribute types:
Partition fields of type UUID must be of binary subtype 4. Atlas skips partition fields of type UUID with subtype 3.
To learn more about the supported partition attribute types, see Partition Attribute Types.
While partitions improve query performance, queries that don't contain these fields require a full collection scan of all archived documents, which will take longer and increase your costs. To learn more about how partitions improve your query performance in Atlas Data Federation, see Data Structure in S3.
You can review the following archiving rule settings:
The name of the database and collection
The name of the date field (for Date Match only)
The number of days to keep data on the Atlas cluster (for Date Match only)
The number of days after which to delete archived data
The frequency and time window for archiving data
The custom query to use to identify data to archive (for Custom Criteria only)
The partition fields
Click Back to edit these settings if needed.
Copy and run the displayed query in your
mongosh shell to see the documents that match the criteria in the rule you defined in step 5.
You can run explain on the query to check whether it uses an index. Proceed to the next step to create the index if the fields are not indexed. If the fields are already indexed, skip to step 11.
(Optional) Copy and run the displayed query in your
mongosh to create the required index. This ensures that your data is indexed for optimal performance.
Click Begin Archiving in the Confirm an online archive tab.
Click Confirm in the Begin Archiving window.
Once your document is queued for archiving, you can no longer edit the document. See Restore Archived Data to move archived data back into the live Atlas cluster.
You can create up to 50 online archives per cluster and up to 20 can be active per cluster. The following limitations apply:
You can configure multiple online archives in the same namespace, but only one can be active at any given time.
You cannot create multiple online archives on the same fields in the same collection.
You can't access your online archive during the following scenarios:
A full outage of the primary region of your cluster.
An outage of AWS S3 where your archived data is stored.