NVIDIA / NVIDIA/nvcf

refactor(pylon): group backend-specific settings into per-backend config

Open
#784 0 comments 0 reactions 1 assignee View on GitHub

@along-2017 is already working on this.

Since Aug 11, 2026.

Dominant language
Go
Stars
218
Forks
72
Avg merge
1d 12h
Merged PRs (30d)
427

Description

Description

Pylon selects its upstream engine dialect with one flag, --pylon-upstream-backend <dynamo|passthrough>. Backend-specific settings are flat flags today; the priority ceiling (--pylon-priority-ceiling) applies to the Dynamo mapping but is not grouped with the backend choice.

When a second real backend lands, flat flags stop scaling: each backend brings its own settings, and unrelated flags would apply to backends that ignore them. Introduce a per-backend config block (one section per backend, only the active one read) so settings live with the dialect they configure.

Definition of Done

  • Backend-specific settings are grouped per backend; a setting for one backend cannot silently apply to another.
  • The existing flags keep working or have a documented migration.
  • docs/api-gateway-contract.md reflects the config surface.

Resources

  • Backend enum and Dynamo module: src/libraries/rust/stargate/crates/pylon-lib/src/quic_http_tunnel/backend.rs
  • Flags: src/libraries/rust/stargate/crates/pylon/src/main.rs

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Assessment

This issue has not been assessed yet.

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.