Metrics, traces, logs and everything they cost.
7 items
A Hacker News post said Claude Code burns 15 to 20 percent of tokens on a status line. The status line is innocent. I reproduced the real culprit, an agent summary, and measured what it costs.
I sent 200 requests across eleven kinds of hand-off with the OpenTelemetry Python SDK. Five lost every trace, batching lost 90%, and the failures you most want to trace were the ones left behind.
I simulated Google SRE Workbook burn-rate paging rules on a 99.9% SLO. They bound the budget lost before a page, but below about 900 requests a day they page for nearly every error.
Prometheus reports fractional counts for counters that only ever increment by one. The reason is a specific extrapolation rule, and knowing it tells you exactly when the number can be trusted.
Cardinality does not creep up on you — it multiplies. Here is the arithmetic behind an exploding metrics bill, the four places the multiplication hides, and what actually shrinks it.
A beginner-friendly SRE guide to service level indicators, objectives and agreements, with a comparison table, the downtime each "nine" allows, and a real cloud SLA read line by line.
What observability actually means, how metrics, logs and traces each answer different questions, how a trace is built, where OpenTelemetry fits, and how SLOs turn signals into alerts.