I'm building a Kubernetes app and want traces, logs, and metrics in one place

A good way to get traces, logs, and metrics in one place for a Kubernetes app is to use an observability stack built around:

• Metrics: Prometheus + Grafana
• Logs: Loki + Promtail/Fluent Bit
• Traces: Jaeger or Tempo
• Correlation / UI: Grafana

Recommended architecture

Option 1: Grafana-centric stack

This is the most common “one place” setup.

• Grafana: single UI for dashboards, logs, and traces
• Prometheus: collects Kubernetes/app metrics
• Loki: stores logs
• Tempo or Jaeger: stores traces
• OpenTelemetry Collector: standard way to receive and route telemetry
• Promtail or Fluent Bit: ships container logs to Loki

Why this is a good choice

• One dashboard for all telemetry
• Easy correlation between logs, traces, and metrics
• Kubernetes-friendly
• Supports OpenTelemetry, which is becoming the standard

Best-practice setup

If you’re starting fresh, I’d suggest:

1. Instrument your app with OpenTelemetry

   • Use OTel SDKs for traces/metrics
   • Propagate trace IDs into logs

2. Run OpenTelemetry Collector in-cluster

   • Receives telemetry from your app
   • Exports:
     • metrics to Prometheus
     • traces to Tempo/Jaeger
     • logs to Loki if you’re sending app logs through OTel, or use Fluent Bit/Promtail for stdout logs

3. Use Grafana as the central UI

   • Explore logs, traces, and metrics together
   • Link trace IDs from logs to traces

Typical stack choices

For a simple setup

• Grafana
• Prometheus
• Loki
• Tempo
• OpenTelemetry Collector

If you want the classic tracing backend

• Grafana
• Prometheus
• Loki
• Jaeger
• OpenTelemetry Collector

If you want the simplest operational path

Consider Grafana Cloud or another managed observability platform:

• less cluster maintenance
• unified UI
• scales better with less setup effort

Important implementation detail

To really get “one place,” make sure your logs include a trace_id and ideally span_id. That lets you jump from:

• a log line → trace
• a trace → related logs and metrics

Example Kubernetes flow

• App emits OpenTelemetry traces/metrics
• App logs go to stdout with trace context fields
• Fluent Bit ships logs to Loki
• OTel Collector ships traces to Tempo and metrics to Prometheus
• Grafana queries all of it

If you want, I can give you:

1. a recommended Helm-based Kubernetes setup,
2. a minimal architecture diagram, or
3. a step-by-step implementation plan for your app.