View Source Cachex.Router.Jump (Cachex v4.0.0)

Routing implementation based on Jump Consistent Hash.

This implementation backed Cachex's distribution in the v3.x lineage, and is suitable for clusters of a static size. Each key is hashed and then slotted against a node in the cluster. Please note that the hash algorithm should not be relied upon and is not considered part of the public API.

The initialization of this router accepts a :nodes option which enables the user to define the nodes to route amongst. If this is not provided the router will default to detecting a cluster via Node.self/0 and Node.list/1.

For more information on the algorithm backing this router, please see the appropriate publication.

Summary

Functions

init(cache, options \\ [])

Initialize a jump hash routing state for a cache.

nodes(nodes)

Retrieve the list of nodes from a jump hash routing state.

route(nodes, key)

Route a key to a node in a jump hash routing state.

Functions

Link to this function

init(cache, options \\ [])

View Source 
@spec init(cache :: Cachex.t(), options :: Keyword.t()) :: [atom()]

Initialize a jump hash routing state for a cache.

Options

  • :nodes

    The :nodes option allows a user to provide a list of nodes to treat as a cluster. If this is not provided, the cluster will be inferred by using Node.self/0 and Node.list/1.

Link to this function

nodes(nodes)

View Source 
@spec nodes(nodes :: [atom()]) :: [atom()]

Retrieve the list of nodes from a jump hash routing state.

Link to this function

route(nodes, key)

View Source 
@spec route(nodes :: [atom()], key :: any()) :: atom()

Route a key to a node in a jump hash routing state.