Observability Setup

Define metrics, logs, traces, dashboards and alerts for a service

Shape your prompt

7 fields

Service namerequired

What does the service do & what matters?required

Observability pillars

Tooling stack

Define SLOs & error budgetsInclude alerting rules

Constraints

Your prompt

1,013 characters

The raw prompt, unchanged.

Still needed: Service name, What does the service do & what matters? — the preview updates as you type.

Output23 lines · 1,013 chars

You are an observability/SRE engineer. Design an observability setup for the service "".

## Service

- Tooling: OpenTelemetry + vendor backend
- Pillars to instrument: Metrics, Logs, Traces

## Instrumentation plan
- Define the golden signals (latency, traffic, errors, saturation) and the specific metrics, log fields and span attributes to emit.
- Use consistent naming and bounded label/tag cardinality.
- Structured, correlatable logs (trace/span IDs propagated).
- Define SLIs, SLOs and error budgets tied to user-facing impact.
- Author alerting rules that page on symptoms (SLO burn), not noisy causes, with runbook links.

## Dashboards
- A service overview dashboard and a deep-dive/troubleshooting dashboard.

## Deliverables
1. The instrumentation spec and config (collector/exporter/scrape config as code).
2. Dashboard definitions and alert rules as code.
3. A short rationale linking each signal to a failure mode.

Use production-grade defaults and explain notable trade-offs; ask only if blocked.