Home › Key concepts › Observability: logs, metrics, traces
Key concept · Walk (intermediate) · the same on every platform
Observability: logs, metrics, traces
See what your system is doing from the outside, especially when it goes wrong.
What it is
Logs are records of events. Metrics are numbers over time, such as requests per second or error rate. Traces follow one request across every service and step. OpenTelemetry is the open standard for collecting all three.
Why it matters
You cannot fix what you cannot see. Good traces turn an hour of guessing into a minute of reading.
An example
One trace shows a slow reply spent 6 of 7 seconds waiting on a search call.
The common mistake
Logging everything as free text and having no way to follow one request.
Where you meet it on each platform
The idea stays the same; only the product name changes.
| Job | AWS | Azure | Google Cloud | Databricks | Open source |
|---|---|---|---|---|---|
| Monitoringlogs, metrics and traces, so you see trouble first | CloudWatch + X-Ray | Azure Monitor + App Insights | Cloud Logging + Trace | System tables + MLflow tracing | OpenTelemetry + Grafana |
| Tracingsee every model call, tool call, token and cent | CloudWatch + X-Ray (OpenTelemetry) | App Insights (OpenTelemetry) | Cloud Trace (OpenTelemetry) | MLflow tracing | Phoenix or Langfuse |
Related ideas
Where it shows up
Part of the Walk level on these paths: ML Engineering (production ml) · Software Engineering (running services).