Introduction to buckets
Buckets are object containers that are used to control access to objects and to set properties that define attributes for all contained objects, such as retention periods and quotas.
In S3, object containers are called buckets and this term has been adopted as a general term in ECS. In Atmos, the equivalent of a bucket is a subtenant. In Swift, the equivalent of a bucket is a container. In CAS, a bucket is a CAS pool.
In ECS, buckets are assigned a type, which can be S3, Swift, Atmos, or CAS. S3, Atmos, or Swift buckets can be configured to support file system access (for NFS and HDFS). A bucket that is configured for file system access can be read and written by using its object protocol and by using the NFS or HDFS protocol. S3 and Swift buckets can also be accessed using each other's protocol. Accessing a bucket using more than one protocol is often referred to as cross-head support.
You can create buckets for each object protocol using its API, usually using a client that supports the appropriate protocol. You can also create S3, file system-enabled (NFS or HDFS), and CAS buckets using the ECS Portal and the ECS Management REST API.
You can create buckets for each object protocol using its API, usually using a client that supports the appropriate protocol. For information about how to create a bucket using the S3 API, click here.
You can also create S3, file system-enabled (NFS or HDFS), and CAS buckets using the ECS Portal and the ECS Management REST API.
Bucket names for the different object protocols must conform to the ECS specifications described in Bucket and key naming conventions.
Bucket ownership
A bucket is assigned to a namespace and object users are also assigned to a namespace. An object user can create buckets only in the namespace to which the object user is assigned. An ECS System or Namespace Administrator can assign the object user as the owner of a bucket, or a grantee in a bucket ACL, even if the user does not belong to the same namespace as the bucket, so that buckets can be shared between users in different namespaces. For example, in an organization where a namespace is a department, a bucket can be shared between users in different departments.
Bucket access
Objects in a bucket that belong to a replication group spanning multiple VDCs can be accessed from all of the VDCs in the replication group. Objects in a bucket that belongs to a replication group that is associated with only one VDC can be accessed from only that VDC. Buckets cannot be accessed or listed from other VDCs that are not in the replication group. However, because the identity of a bucket and its metadata, such as its ACL, are global management information in ECS, and the global management information is replicated across the system storage pools, the existence of the bucket can be seen from all VDCs in the federation.
For information about how objects in buckets can be accessed during site outages, see TSO behavior with the ADO bucket setting turned on.
In the ECS Portal, you can: