Home > Storage > ObjectScale and ECS > Product Documentation > ECS: Overview and Architecture > Single site deployment
During a single site, or single cluster initial deployment, nodes are first added into a SP. SPs are logical containers of physical nodes. SP configuration involves selecting the required minimum number of available nodes and choosing the default 12+4 or cold archive 10+2 EC scheme. Critical alert levels may be set during SP configuration initially and in the future; however, EC schema cannot be changed after SP initialization. The first SP created is designated as the system SP and is used to store system metadata. The system SP cannot be deleted.
Clusters generally contain one or two SPs, as shown in the following figure—one for each EC schema; however, if an organization requires physical separation of data, additional SPs are used to implement boundaries.
After the initialization of the first SP, a VDC can be created. VDC configuration involves designating replication and management endpoints. Although system SP initialization is required prior to VDC creation, VDC configuration does not assign SPs but rather the IP addresses of nodes.
After a VDC is created, RGs are configured. RGs are global resources with configuration involving designating at least one VDC, itself, in the single- or initial-site setup, along with one of the VDC’s SPs. An RG with a single VDC member protects data locally at the disk, node, and rack level. The next section expands on RGs to include multisite deployments.
Namespaces are global resources created and assigned to an RG. At the namespace level retention policies, quotas, compliance, and namespace administrators are defined. Access During Outage (ADO) can be configured at the namespace level which is covered in the next section. Generally, it is at the namespace level where tenants are organized. Tenants may be an application instance or team, user, business group, or any other grouping that makes sense to the organization.
Buckets are global resources that can span multiple sites. Bucket creation involves assigning it to a namespace and an RG. The bucket level is where ownership and file or CAS access is enabled. The following figure shows one SP in a VDC with a namespace containing two buckets: