Namespaces

Namespaces let you run multiple independent configurations within a single Coyote cluster.

For example, you can have two cache namespaces one with 4GB capacity that’s backed by disk, and another with 1GB capacity and backed by RAM; without needing separate clusters.

Creating namespaces

A namespace has to be created before used, and each module has its own namespace API

coyote.cache.namespace.create
coyote.kv.namespace.create
...

The create API is an upsert it creates the namespace if it doesn’t exist, or updates it if it does. Some properties are immutable after creation; attempting to change them returns an error.

All configuration values are optional and fall back to the default namespace’s values if unset.

Namespaces are scoped per module type. A cache namespace named payments is independent from a kv namespace named payments. Using a namespace that doesn’t exist returns an error.

Using namespaces

A namespace is identified by a prefix on your key (e.g. a cache key and stream topic), separated by a colon:

foo/bar          # uses the default namespace
payments:foo/bar # uses the "payments" namespace

Keys without a prefix belong to the default namespace for each module.

Configuration

Each namespace can be configured with:

As well as additional module specific configuration options. For the full set of configurations, please refer to each module’s namespace.create function.

Bootstrapping namespaces

Coyote supports automatically running a bootstrap script that lets you automatically create and configure you wanted namespaces.

The bootstrapping processes runs on every node, so non-immutable settings stay in sync with your configuration.

Default namespace

Every module type has a default namespace with sensible defaults. You can either create them yourself, with your own configurations, or let Coyote create them for you.