GitHub Releases Docs: ntsd.dev/docs/2.4 Discuss: #ntsd on Libera.Chat Have you tried NTSD 2.4 yet? Let me know your use case in the comments!
If you’re running high-traffic Nginx (or general Linux) infrastructure, you know the struggle: balancing granular real-time metrics with system performance. Enter — the latest iteration of the Nginx Tiny Stats Daemon — and it’s a quiet game-changer. ntsd 2.4
For those unfamiliar, NTSD is a lightweight, event-driven statistics collector. Unlike Prometheus + exporters or custom Lua scripts, NTSD sits in the background, consuming minimal RAM and CPU, yet delivers JSON endpoints, log parsing, and now, even deeper insights. 1. Reduced Memory Footprint (~30% less) The 2.4 release refactors internal hash maps and connection pooling. On a server handling 10k+ requests/second, memory usage dropped from ~18MB to ~12MB. For containerized deployments (Docker, k8s), that’s noticeable. 2. Structured Logging with auto-rotate NTSD now natively supports JSON-formatted logs with log rotation signals. No more tail -f grepping — pipe directly into loki or fluentd . You can enable it via: GitHub Releases Docs: ntsd
NTSD 2.4 Released: Smarter Metrics, Lower Overhead for High-Traffic Servers Enter — the latest iteration of the Nginx