Bucket in Object Storage
A bucket is a logical entity for storing objects.
Naming buckets
A bucket name is used as part of the data access URL and is visible to your users. For example, it may look like https://storage.ai.nebius.cloud/bucket-name
.
The naming requirements are as follows:
-
Bucket names are unique throughout Object Storage, which means you cannot create two buckets with the same name, even in different folders belonging to different clouds. You should keep this in mind if you are going to create buckets automatically through the API.
-
Bucket names are subject to the following restrictions:
- The name must be from 3 to 63 characters long.
- The name may contain lowercase Latin letters, numbers, hyphens, and periods.
- The first and last characters must be letters or numbers.
- The characters to the right and left of the period must be letters or numbers.
- The name must not look like an IP address, e.g.,
10.1.3.9
.
Bucket URL
You can use the following URL formats to access a bucket:
https://storage.ai.nebius.cloud/<bucket>?<parameters>
http://<bucket>.storage.ai.nebius.cloud?<parameters>
Note
If you use AWS SDK for .NET, access the s3.ai.nebius.cloud
host.
Bucket settings
You can:
-
Limit the maximum bucket size.
Object Storage will not allow you to upload an object if doing so leads to exceeding the maximum bucket size.
-
Set up object lifecycles.
Bucket access
By default, buckets are created with restricted access, as specified in the IAM settings. You can use the management console to enable public access:
- To read objects in a bucket.
- To access the list of objects in a bucket.
- To read settings:
Public access to each operation is provided separately. This means, if you have opened read-only access to your objects, anonymous users cannot get the list of objects and bucket settings.
Recommendations and limitations
-
You cannot rename buckets.
-
The number of buckets does not affect the performance of Object Storage. How many buckets you use to store your data is up to you.
-
Buckets cannot be nested.
-
In the management console, information about the number of objects in a bucket and the used space is updated with a few minutes' delay.
-
You can only delete an empty bucket.
-
After deleting a bucket, you may not be able to immediately create a new bucket with the same name. There is also a risk that another Nebius AI user might create a bucket with the name you released before you claim it again. Do not delete buckets without a reason.
-
Uploading objects quickly one by one may lead to exceeding the specified maximum bucket size.
-
After deleting objects from a bucket, their storage capacity is retained for some time.
Note
If you limit the maximum size of a bucket, it may remain temporarily unavailable for writes even after you free up enough space for new objects.