multilevel

Available in: Cloud, Self-Managed

Combines multiple caches as levels, performing read-through and write-through operations across them.

caches:
  label: ""
  multilevel:
    - label: ""
      memory:
        default_ttl: 5m
        compaction_interval: 60s
    - label: ""
      redis:
        url: redis://localhost:6379
        expiration: 24h

Examples

Hot and cold cache

The multilevel cache is useful for reducing traffic against a remote cache by routing it through a local cache. In the following example requests will only go through to the memcached server if the local memory cache is missing the key.

pipeline:
  processors:
    - branch:
        processors:
          - cache:
              resource: leveled
              operator: get
              key: ${! json("key") }
          - catch:
            - mapping: 'root = {"err":error()}'
        result_map: 'root.result = this'

cache_resources:
  - label: leveled
    multilevel: [ hot, cold ]

  - label: hot
    memory:
      default_ttl: 60s

  - label: cold
    memcached:
      addresses: [ TODO:11211 ]
      default_ttl: 60s