Ross ROSS = Recommend OSS · open-source software intelligence for agents

parca-dev/parca

Continuous profiling for analysis of CPU and memory usage, down to the line number and throughout time. Saving infrastructure cost, improving performance, and increasing reliability. observed · 2026-08-28

github.com/parca-dev/parca · homepage · TypeScript · Apache-2.0 (permissive) observed · 2026-08-28

Health v2 · maintenance only

89/100

  • Activity 99
  • Release rhythm 71
  • Longevity 100
How is this computed?

round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10) — computed 2026-09-03. Adoption (stars, forks) is never an input.

  • gap_med: 47
  • age_days: 2836
  • days_rel: 118
  • days_push: 7
  • n_releases_24m: 10

Full methodology

Adoption not part of the score

4951 stars · 255 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

Parca is an open-source continuous profiling server that collects, stores, and queries CPU, memory, and I/O profiles over time, down to the line number. It pairs with an eBPF-based agent that automatically discovers and profiles targets across infrastructure with near-zero instrumentation and low overhead.

Use cases

  • continuously profile CPU and memory usage in production
  • find hot code paths to optimize and reduce infrastructure cost
  • compare profiling data across deploys, versions, or regions
  • investigate memory leaks and CPU spikes during incidents
  • auto-discover and profile all workloads in a Kubernetes cluster
  • store and query pprof profiles with a label-based query language

When to choose

  • you want always-on, zero-instrumentation profiling across C, C++, Rust, Go, Python, Ruby, PHP, Node.js, and more
  • you run Kubernetes or systemd-based infrastructure and want automatic target discovery via eBPF
  • you need to store, aggregate, and compare pprof profiles over time with a web UI and flame graphs
  • you want a self-hosted, open-source alternative to commercial continuous profilers

When to avoid

  • you need tracing or distributed request-level observability rather than sampling profiles
  • you cannot run privileged/root eBPF programs on your Linux hosts
  • you need profiling on non-Linux platforms, since the eBPF agent is Linux-only
  • you only need ad-hoc, one-off profiling of a single process

Facets

application · maturity active

monitoring tracing analytics search-engine self-hosted monitoring performance cloud-computing self-hosted go continuous-profiling pprof ebpf observability flamegraph prometheus cpu-profiling memory-profiling devops containers linux docker kubernetes web-server

8 sources

Member repositories

RepositoryRoleHealth v2
parca-dev/parcamain89

For agents

markdown · JSON · MCP: product_card(name="parca-dev/parca")

Data as of 2026-08-30T08:39:29.467469+00:00 · Report a problem