continuous-profiling 开源项目推荐（26 个）

grafana /pyroscope

持续分析平台。将性能问题压缩到一行代码中

parca-dev /parca

持续性能分析，用于分析 CPU 和内存使用情况，可精确到行号并贯穿整个时间范围。节省基础设施成本、提升性能并提高可靠性。

yandex /perforator

Perforator 是一款面向大型数据中心的集群级持续性能分析工具

grafana-cold-storage /phlare

可水平扩展、高可用、多租户的持续性能分析聚合系统

parca-dev /parca-agent

基于 eBPF 的常开 CPU/GPU 分析器，可自动发现 Kubernetes 和 systemd 中的目标，无需修改代码或重启！

profefe /profefe

用于长期事后分析的持续性能分析

polarsignals /rust-jemalloc-pprof

Convert jemalloc heap profiles to pprof to understand memory usage, fix memory leaks, and fix OOM Kills.

xyctruth /profiler

Continuous profiling based on pprof

polarsignals /pprofme

Easily share pprof formatted profiles from your terminal.

Query-farm /pyroscope

DuckDB Pyroscope Extension for Continuous Profiling

everythings-gonna-be-alright /phpScope

PHP profiler that sends CPU sampling data to Pyroscope server.

mohammed90 /caddy_profiling

Continuous profiling app for Caddy

molesignal /molesignal

Open-source, fully self-hostable observability and incident response. Telemetry, alerting, on-call, status pages and AI investigation in one system.

ricoberger /parca-operator

Deploy and Operator Parca

parca-dev /jsonnet-libs

Experimental Jsonnet libraries for Parca

polarsignals /docs

Polar Signals Cloud Documentation

kubehero-io /platform

一次安装即可实现 Kubernetes 成本分摊、日志、持续剖析和 eBPF 网络成本。提供兼容 OpenCost 和 Loki 的 API、通过受保护的 CRD 应用的资源合理配置，以及带 MCP 服务器的只读代理。

eugeneliukindev /fastapi-observatory

FastAPI under a full observability stack: metrics, logs, traces, profiles and browser RUM via OpenTelemetry, Alloy and the Grafana LGTM stack, in Docker Compose or Kubernetes.

tracewick /tracewick

Sampling profiler for PHP built on ext-excimer: flame graphs, pprof and continuous profiling exports (alpha).

nubenetes /observability-platforms-comparison

Enterprise Hybrid Cloud & OpenShift Observability Platforms Comparison: Exhaustive architectural evaluation of 12 platforms across the 7 Pillars of Observability in Air-Gapped and Regulated Environments.