Skip to content

docs(blog): clarify HTTP/3 proxy rationale in one-core bottleneck article#207

Merged
josecelano merged 1 commit into
developfrom
198-clarify-http3-proxy-rationale
Jun 15, 2026
Merged

docs(blog): clarify HTTP/3 proxy rationale in one-core bottleneck article#207
josecelano merged 1 commit into
developfrom
198-clarify-http3-proxy-rationale

Conversation

@josecelano

Copy link
Copy Markdown
Member

Summary

Clarify that disabling HTTP/3 was not the root-cause fix for the one-core softirq bottleneck, and that RPS/RFS packet steering was the actual remediation. Also add protocol boundary notes and link follow-up issues for traceability.

Changes

  • Add summary Callout near top with clarification, protocol boundary note, and follow-up links
  • Enhance Phase 2 conclusion to explicitly state QUIC was not the driver
  • Add HTTP/3 as a separate product/operations decision in Final Operational Decision
  • Add follow-up references for torrust-tracker-demo#31 and torrust-tracker#1736

Related issue

Closes #198

Checklist

  • npm run lint passes
  • npm run check passes
  • Manually tested in the browser (the article renders correctly with all changes)

…icle

- Add summary Callout near top clarifying that disabling HTTP/3 was not the fix
- Add protocol boundary note (client->proxy HTTP/3, proxy->backend HTTP)
- Enhance Phase 2 conclusion to explicitly state QUIC was not the driver
- Add HTTP/3 as a separate product/ops decision in Final Operational Decision
- Add follow-up references for torrust-tracker-demo#31 and torrust-tracker#1736

Closes #198
Copilot AI review requested due to automatic review settings June 15, 2026 12:08

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.

Pull request overview

This PR updates the “one-core softirq bottleneck” blog article to clearly separate the Phase 2 HTTP/3 (QUIC) disablement experiment from the actual remediation (enabling RPS/RFS), and adds protocol-boundary clarification plus follow-up issue links for traceability.

Changes:

  • Add an early warning callout explicitly stating HTTP/3 disablement was not the root-cause fix, and that RPS/RFS packet steering was.
  • Strengthen the Phase 2 conclusion to state QUIC on UDP 443 was not driving the hotspot, and frame HTTP/3 as an ops/product decision.
  • Add follow-up issue references (torrust-tracker-demo#31, torrust-tracker#1736) in the operational decision and references sections.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

@josecelano josecelano self-assigned this Jun 15, 2026
@josecelano

Copy link
Copy Markdown
Member Author

ACK 99c0d15

@josecelano josecelano merged commit dbdb7cf into develop Jun 15, 2026
2 checks passed
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.

docs(blog): clarify HTTP/3 proxy rationale in one-core bottleneck article

2 participants