Skip to content

docs(docker): note /crawl result order, and that app.workers is not read - #2292

Open
talelboussetta wants to merge 2 commits into
unclecode:developfrom
talelboussetta:docs/docker-result-order-and-workers
Open

talelboussetta wants to merge 2 commits into
unclecode:developfrom
talelboussetta:docs/docker-result-order-and-workers

Conversation

@talelboussetta

@talelboussetta talelboussetta commented Sep 25, 2026 •

Copy link
Copy Markdown

Summary

Two small corrections to the Docker server docs.

  1. /crawl result order. results are not in the order of urls: MemoryAdaptiveDispatcher.run_urls appends each result as asyncio.wait(..., FIRST_COMPLETED) returns its task, so a slow page comes back after a fast one, and pages that finish together come back in no set order. Nothing in the Docker README or the self-hosting guide says so, and indexing results[i] against urls[i] files one page under another page's URL with no error. A note after the Simple Crawl example, in both, says to match by url.
  2. app.workers in config.yml is not read. supervisord.conf starts gunicorn with a fixed --workers 1, and server.py's direct-run uvicorn.run(...) does not pass it. Marked the way the README already marks the direct-run-only port.

List of files changed and why

  • deploy/docker/README.md, docs/md_v2/core/self-hosting.md: the result-order note.
  • deploy/docker/config.yml: a comment on app.workers.

How Has This Been Tested?

Docs and a YAML comment only; config.yml still loads, and the offline suites are unchanged against develop. The order, on an image built from develop @ 1f68e5b, against a local test site, so CRAWL4AI_ALLOW_INTERNAL_URLS=true: two URLs, the first held back 3 s by its own per-URL config:

[order] sent ['/index.html', '/a.html'] -> HTTP 200, results ['/a.html', '/index.html']

cc @ntohidi

Checklist:

  • My code follows the style guidelines of this project
  • I have performed a self-review of my own code
  • I have commented my code, particularly in hard-to-understand areas
  • I have made corresponding changes to the documentation
  • I have added/updated unit tests that prove my fix is effective or that my feature works
  • New and existing unit tests pass locally with my changes

/crawl returns results in the order the pages finished:
MemoryAdaptiveDispatcher.run_urls appends each task as it completes. Say so
next to the Simple Crawl example in the Docker README and the self-hosting
guide, and to match results by url rather than by position.

app.workers in config.yml is not read anywhere: supervisord.conf starts
gunicorn with a fixed --workers 1, and server.py's direct-run path does not
pass it. Mark it the way the README already marks the direct-run-only port.
Copilot AI lite review requested due to automatic review settings September 25, 2026 10:19

Copilot AI left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Copilot review overview

🟡 Changes recommended

Scope ordering guidance to regular multi-URL crawls and clarify that ties are unspecified.

Get a fresh assessment by requesting another Copilot review.

Review effort: Lite
Findings: 2 Low severity

Open (2)
What changed in this PR

Updates Docker documentation for /crawl result ordering and the unused app.workers setting.

Changes:

  • Adds URL-based result matching guidance.
  • Documents that app.workers is not read by Docker startup.
File Summary
docs/​md_v2/​core/​self-hosting.md Adds /crawl ordering guidance.
deploy/​docker/​README.md Adds result-order guidance.
deploy/​docker/​config.yml Documents the unused worker setting.

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.

Comment thread deploy/docker/README.md Outdated
Comment thread docs/md_v2/core/self-hosting.md Outdated
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants