Folders and namespaces

A CONNECT account can only be associated with one CONNECT data services tenant. Each tenant may have more than one namespace, and each namespace maps to a unique folder in a CONNECT account.

A namespace represents a logical unit of organization for data within a tenant. When you create a folder in CONNECT and enable the CONNECT data services tile, a namespace is automatically created in CONNECT data services. Multiple folders in CONNECT can have CONNECT data services turned on. Each folder represents a namespace, and all of the namespaces are part of one tenant and one CONNECT account.

Data processing resources are allocated to a namespace after you create a folder and turn on CONNECT data services. For example, SDS and asset services, and the associated storage resources are allocated to support a namespace. Each namespace and its resources are distinct and separate from all other namespaces. For example, you can create an SdsType or an SdsStream object with the same name in two different namespaces.

Identity resources in CONNECT data services, such as users and clients, are scoped globally across namespaces. For example, it is not possible to have User1 in the Production namespace but not in the Staging namespace. In CONNECT, users are globally-scoped with the ability to exist and have permissions across all folders in a CONNECT account.

Data stored within a namespace is tied to the namespace's region. You cannot directly transfer this data to any other namespace (or region). To move data between namespaces, you must export the data from the source namespace in CONNECT data services, then import it into the destination namespace (for more information, see Transfer PI System data to CONNECT data services).

As the administrator or developer setting up folders and namespaces, consider the following important points:

• Before CONNECT data services can receive data for a given tenant, a namespace must exist within the scope of the CONNECT data services tenant.

• A namespace is linked to an associated CONNECT folder.

• The CONNECT folder specifies the data storage region; for example, West US or Northern Europe. Alternatively, a CONNECT data services namespace can have its own defined region separate from its CONNECT folder’s region if CONNECT data services does not support the CONNECT folder’s region.

• The folder name is saved as the namespace name.

• A namespace's name can only be changed by editing the associated CONNECT folder name.

Best practices

We recommend that you create CONNECT data services namespaces for each region that data needs to be stored in to help meet data sovereignty requirements or latency requirements. Another option would be to use CONNECT data services namespaces to segment data for development, staging, and production scenarios. Note that objects (such as streams) and collections (such as Data Views) do not automatically inherit security from the namespace with which they are associated.

PI Server counterpart

A namespace is similar to a full PI Server. Much like a PI Server, a namespace has its own resources and it is not typical or easy to use data from multiple namespaces at the same time. It is reasonable for a tenant to have only one namespace.

Namespace IDs

A namespace ID is defined by the solution ID in the CONNECT folder. You can view a namespace's ID by selecting the Change Namespace dropdown list in the CONNECT data services portal. The namespace name is synchronized with the CONNECT folder name.

Querying data across namespaces

When querying API endpoints, the namespace ID is part of the URL. The API URL takes the form https://{server}/api/{version}/Tenants/{tenant id}/Namespaces/{namespace id}, followed by the API path, such as /Streams. Because the URL contains the namespace, it is not possible to make a single API request for data across multiple namespaces.

Namespace deletion

A namespace gets suspended when the administrator turns off the CONNECT data services tile in a CONNECT folder. The customer then has 30 days to get the namespace reactivated. During the 30-day suspension period, neither reading or writing to the namespace is permitted. After 30 days, the namespace is deleted.