ci(maintain-deploy): also discover LLM360-head PRs open against sgl-project/sglang#10
Merged
DavidBellamy merged 1 commit intollm360-mainfrom Apr 19, 2026
Conversation
…roject/sglang The other 4 LLM360 forks (miles, harbor, smg, Megatron-LM) use their stabilized LLM360-friendly upstream (radixark, harbor-framework, lightseekorg) as BOTH the deploy reset base AND the source of open PRs with LLM360-owned heads. That pattern breaks for sglang: sgl-project/ sglang moves very fast and does NOT contain our patch stack, so resetting deploy to it would wipe sglang-miles/TITO/MTP/rail-map every build. The old self-referential workaround (UPSTREAM=LLM360/sglang) kept the reset correct but left a gap: upstream PRs the team authors against sgl-project/sglang (e.g. rail-map JSON passthrough in sgl-project#23003) don't land in our nightly image until they merge upstream. That block tonight's agentic RL pilots. Decouple the two concerns: - DEPLOY_BASE + DEPLOY_BASE_BRANCH = LLM360/sglang:llm360-main (what we reset to; contains our patch stack) - REAL_UPSTREAM + REAL_UPSTREAM_BRANCH = sgl-project/sglang:main (where we also query for open PRs with LLM360-owned heads) Deploy = DEPLOY_BASE + cross-fork PRs into llm360-main + fork-local PRs + LLM360-head PRs open against sgl-project/sglang:main. Dedup by branch name. Branches are all fetchable via origin=LLM360/sglang since the head branch always lives there regardless of which base the PR points at.
DavidBellamy
added a commit
that referenced
this pull request
Apr 19, 2026
…project#23003 Upstream PR sgl-project#23003 (fix/json-ib-device-passthrough) is superseded by #9 (fix/json-ib-device-passthrough-llm360-main), which mirrors the same 7-line patch on top of llm360-main. The upstream branch was based on sgl-project/sglang:main and line-drifts when merged into llm360-main — was triggering a merge conflict in every octopus run after the PR #10 decouple landed. Port harbor's SKIP_UPSTREAM_BRANCHES pattern: env var + jq `NOT_SKIPPED` filter applied to both UPSTREAM_PR_STATE/BRANCHES (cross-fork PRs on LLM360/sglang:llm360-main) and REAL_UPSTREAM_PR_STATE/BRANCHES (upstream sgl-project/sglang PRs). Removes the skipped branch cleanly from the state fingerprint too, so re-running doesn't see it as a pending change.
DavidBellamy
added a commit
that referenced
this pull request
Apr 19, 2026
…project#23003 (#11) Upstream PR sgl-project#23003 (fix/json-ib-device-passthrough) is superseded by #9 (fix/json-ib-device-passthrough-llm360-main), which mirrors the same 7-line patch on top of llm360-main. The upstream branch was based on sgl-project/sglang:main and line-drifts when merged into llm360-main — was triggering a merge conflict in every octopus run after the PR #10 decouple landed. Port harbor's SKIP_UPSTREAM_BRANCHES pattern: env var + jq `NOT_SKIPPED` filter applied to both UPSTREAM_PR_STATE/BRANCHES (cross-fork PRs on LLM360/sglang:llm360-main) and REAL_UPSTREAM_PR_STATE/BRANCHES (upstream sgl-project/sglang PRs). Removes the skipped branch cleanly from the state fingerprint too, so re-running doesn't see it as a pending change.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Why
The other 4 LLM360 forks (miles, harbor, smg, Megatron-LM) use their stabilized LLM360-friendly upstream (radixark, harbor-framework, lightseekorg) as BOTH the deploy reset base AND the source of open PRs with LLM360-owned heads. That pattern breaks for sglang: `sgl-project/sglang` moves very fast and does not contain our patch stack, so resetting deploy to it would wipe sglang-miles / TITO / MTP / rail-map every build.
The old self-referential workaround (`UPSTREAM=LLM360/sglang`) kept the reset correct but left a gap: upstream PRs the team authors against `sgl-project/sglang` (e.g. rail-map JSON passthrough in sgl-project#23003) don't land in our nightly image until they merge upstream. That blocked tonight's agentic RL pilots.
What this PR does
Decouples the two concerns by replacing `UPSTREAM` with two pairs:
Deploy is built from:
All discovered branches are still fetchable via `origin = LLM360/sglang` because the head branch always lives on the fork regardless of which base the PR targets.
Verification
After merge, the next maintain-deploy run will pick up open sgl-project PRs with LLM360 heads. Currently that set is only:
Which has already been landed manually via #9 to unblock tonight. After this workflow change, equivalent future upstream PRs flow in without manual mirroring.