Database

A Database is a logical cache that runs on a Capacity Pool. It is the logical container for your keys and values. You connect to a Database with a standard Valkey or Redis client through the shared gateway.

Databases and pools

A Database belongs to exactly one Capacity Pool. You can run several Databases on the same pool, and they share the pool's compute and memory. Because they share capacity, Databases on one pool are subject to noisy-neighbor effects from each other. Isolation is guaranteed at the pool boundary, not between Databases on the same pool. See Isolation.

Choose how to group Databases onto pools based on the isolation and capacity you want. Put workloads that must not contend on separate pools. Pack workloads that can share capacity onto one pool to use it efficiently.

Lifecycle

You create a Database by naming it and assigning it to a pool. Creation is instant, because it does not provision infrastructure. The capacity already exists in the pool. Deleting a Database removes it immediately. The underlying capacity is reclaimed in the background and returns to the pool for reuse.

A Database has no mutable settings today, so there is no update operation. You manage capacity by changing the pool, not the Database.

Connecting

To connect, point a Valkey client at the pool's gateway endpoint with your Database's credentials. See Connectivity and the gateway and the control-plane API reference.