Skip to content

Multi-browser pool / horizontal concurrency #44

Description

@fusengine

Today the engine is single-process: runProbe opens+tears down one browser per probe, and B3 caps concurrency within one process. Horizontal scale = launch N instances manually.

Scope (explore first)

  • A bounded pool of warm browser processes reused across probes (crawlee BrowserPool model: retireBrowserAfterPageCount, maxOpenPagesPerBrowser).
  • Decide: in-process pool of contexts vs multi-process workers.
  • Must preserve per-probe identity/proxy isolation and stealth.

Acceptance

  • Sustained throughput higher than launch-per-probe without unbounded memory.
  • Opt-in; default behavior unchanged.

Risk — touches the core probe lifecycle; needs careful design + benchmarks. Not a quick win.

Activity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions