Production Observability Setup
Prometheus & Grafana Implementation
A self-hosted stack that stays affordable as cardinality grows.
- Turnaround
- 1–2 week sprint
- Who it is for
- Teams wanting observability without a per-host vendor bill.
01 — The problem
Prometheus is easy to start and hard to keep. Cardinality grows, retention becomes expensive, and the dashboards drift from the services they describe.
02 — What is included
Exactly what you get, and nothing vague.
- Prometheus with remote write and a retention strategy that will not surprise you
- Golden-signal dashboards per service: latency, traffic, errors, saturation
- Cardinality review — the thing that makes Prometheus expensive later
- Alertmanager routing with sane grouping and inhibition rules
- Grafana provisioned as code, so dashboards survive the person who made them
03 — What this usually finds
The recurring ones, named up front.
- Labels with unbounded cardinality, such as user or request IDs
- Dashboards for services that no longer exist
- Alerts with no runbook and no owner
- Retention set to a default nobody chose
Listed because they recur, not because they are guaranteed. The audit reports what is actually there.
Next step
Twenty minutes to find out whether this is worth doing.
Every engagement runs under a signed mutual NDA with a read-only IAM role. If the audit would not pay for itself, I will say so on the call.