- Implementation

The steps and options for deploying the with Redis are outlined in the following sections.

Procedure

  1. Implement monitoring
    Although the use of monitoring with Prometheus and Grafana is not mandatory, its use is critical for tuning, and ensuring correct functioning of the cache in production. The defines a comprehensive set of monitoring metrics, which provides detailed insight into cache operation and performance. These metrics can be consumed in Prometheus and displayed with the supplied Grafana dashboards. It is also possible to integrate 3rd-party monitoring to consume the Prometheus-style metrics, but that requires additional configuration.
  2. Select a Redis configuration
    Redis can be installed in a variety of configurations, depending on your performance and high-availability requirements. Alternatives include using the Bitnami Charts to install within the Kubernetes cluster, using Redis Enterprise by RedisLabs, or using Redis as-a-service from a cloud provider. Redis can be installed standalone, or clustered with replicas. Review the Redis Server guide for details.
    This is also a good time to familiarize yourself with Redis and its use by the . See Memory management and HCL Cache in Redis for details.
  3. Configure the Redis client in
    The Redis client configuration in must be updated to match your Redis server setup. The Redis client is configured with a YAML file that contains information about the topology (standalone, cluster, etc), the Redis end-points, TLS and authentication options, and time-outs and thread pool options. The Redis client can be configured in values.yaml and it is stored in a Kubernetes config map. See Redis Client Config for details.
  4. Caching configurations for custom caches
    As the is implemented as a DynaCache Cache Provider, custom caches are enabled for . Caches, by default, enable local and remote caching. Depending on your requirements, you can re-configure custom caches to be local-only, or remote-only. See Cache Configurations and Custom Caching for details.
  5. Cache Tuning
    Large implementations can benefit from additional cache tuning configurations. This can be done with the support of Prometheus and Grafana monitoring. Large single caches, such as baseCache, can be sharded so they can be distributed among multiple Redis masters. Compression and inactivity options are also available to reduce the memory footprint of each cache. See Remote Cache Tuning Configurations for additional details.