Release notes in S3

2026

June

added the ability to manage object and bucket locking via the Control Panel; learn more in the Object Lock and Managing object locking instructions;
added a new multi-AZ pool for data storage — ru-6;
added new events to audit logs: s3.bucket_object_lock.enable and s3.bucket_object_lock.update;
added the ability to connect S3 to other products over a private network via a global router, learn more in the Connect S3 to other products instruction.

April

expanded CopyObject operations in conditional requests.

Now you can use the If-Match and If-None-Match headers in these operations. We have added the corresponding keys to access policies: s3:if-match and s3:if-none-match.

March

added S3 support to audit logs — s3 service;
added the ability to configure Object Lock and lock objects in a bucket;
added new roles s3.admin and s3.user.

Their capabilities are identical to those of the object_storage:admin and object_storage_user roles. You cannot assign the object_storage:admin and object_storage_user roles to new users. Users who have already been assigned these roles will continue to work.

December

added consumption details by bucket.

Starting December 1, 2025, you can view the consumption of each bucket; learn more in the View consumption subsection of the S3 payment model and prices instruction;
added the ability to import an access policy from another bucket.

Policies can be imported from other regions and pools. Rules are migrated in full, including the list of users, actions, conditions, and resources. For resources, the imported bucket name is automatically replaced with the name of the target bucket.

November

added a new storage class — Glacier.

Glacier storage is best suited for long-term data storage. Glacier storage is available only in the ru-3 pool.
You can check the usage cost at my.selectel.ru. Read more about glacier storage in the Storage Classes subsection of the S3 Payment Model and Prices;
added a new pool for data storage — ru-3.