#opentelemetry
4 posts.
How Much Telemetry Should We Keep? Balancing Collection Cost and Diagnostic Value
Examine metric, log, and trace costs from instrumentation to storage, then use aggregation, sampling, label budgets, and retention policies to preserve useful diagnostic evidence.
Where Is the Web Application Slow? A Guide to Narrowing the Problem
Map the path from browser through network, application, and database, then compare evidence from the same user action to narrow a performance problem and verify the fix.
Should Metrics Be Pulled or Pushed? Comparing Collection Models
Compare Prometheus-style scraping and OTLP export across convenience, scalability, failure handling, Kubernetes fit, and current hybrid architecture patterns.
Trace Should Not Break When Kafka Is in the Middle
When Kafka sits inside an MSA flow, tracing does not continue automatically like synchronous HTTP calls. This post explains how to propagate trace context through Kafka headers and how to think about producer and consumer instrumentation.