// the find
oliver006/redis_exporter
Prometheus Exporter for Valkey & Redis Metrics. Supports Valkey 9.x, 8.x, 7.x and various Redis versions
A Prometheus exporter for Redis and Valkey that exposes the full `INFO` output plus key-level metrics, stream metrics, and Sentinel/Cluster topology as Prometheus gauges. It's the de facto standard for this job — if you're running Redis in production and using Prometheus, you're probably already using this. Covers single instances, Sentinel setups, and clusters.
Supports the multi-target `/scrape` pattern correctly, meaning one exporter process can cover many Redis instances without the cardinality explosion that comes from running separate exporters per host. The Lua script hook (`--script`) lets you push custom application-level metrics through the same pipeline without writing a separate exporter. Cluster node autodiscovery via `/discover-cluster-nodes` with HTTP SD is a genuine time-saver in Kubernetes. The password file support uses a JSON map of host→password, so you can handle heterogeneous auth across a fleet without spawning multiple processes.
`--check-keys` with glob patterns runs SCAN on every scrape interval, which can visibly impact a loaded Redis instance — the documentation warns about this but the footgun is real and easy to hit in staging where key counts look harmless. One password per scrape target is the hard architectural limit when using multi-target mode; if your fleet has mixed credentials you're back to running multiple exporter instances. The contrib Grafana dashboards haven't been meaningfully updated in years and don't cover Valkey-specific metrics or the newer stream consumer group fields that the exporter actually exports. No built-in alerting rules beyond the mixin, which is jsonnet-only and adds toolchain friction for teams not already in that ecosystem.