feature: support X-Data-Parallel-Rank header to specific dp-rank.

[doujiang24]

Motivation

From #19268
Added support for X-Data-Parallel-Rank HTTP Header, It could be more friendly for external router.

• Align with vLLM's implementation for external DP routing
• Allow external load balancers/routers to specify DP rank via HTTP header with higher priority than request body

Modifications

• python/sglang/srt/entrypoints/openai/serving_base.py: Added extract_routed_dp_rank_from_header() method
  • Priority: Header > Body routed_dp_rank
  • Validation: Returns 400 Bad Request for non-integer values
• python/sglang/srt/entrypoints/openai/serving_chat.py: Integrated header extraction in _convert_to_internal_request() to use the effective DP rank
• python/sglang/srt/entrypoints/openai/serving_completions.py: Same integration for completion requests

Accuracy Tests

Benchmarking and Profiling

Checklist

• Format your code according to the Format code with pre-commit.
• Add unit tests according to the Run and add unit tests.
• Update documentation according to Write documentations.
• Provide accuracy and speed benchmark results according to Test the accuracy and Benchmark the speed.
• Follow the SGLang code style guidance.

Review Process

1. Ping Merge Oncalls to start the PR flow. See the PR Merge Process.
2. Get approvals from CODEOWNERS and other reviewers.
3. Trigger CI tests with comments or contact authorized users to do so.
   • /tag-run-ci-label, /rerun-failed-ci, /tag-and-rerun-ci
4. After green CI and required approvals, ask Merge Oncalls to merge.

[gemini-code-assist]

Warning

You have reached your daily quota limit. Please wait up to 24 hours and I will start processing your requests again!

[doujiang24]

@hnyls2002 Could you please take a look? Thanks.

[hnyls2002]

/tag-and-rerun-ci

[hnyls2002]

/rerun-stage stage-c-test-8-gpu-h20

[github-actions]

✅ Triggered stage-c-test-8-gpu-h20 to run independently (skipping dependencies).

[github-actions]

🔗 View workflow run

[hnyls2002]

/rerun-stage stage-c-test-8-gpu-h20

[github-actions]

✅ Triggered stage-c-test-8-gpu-h20 to run independently (skipping dependencies).

[github-actions]

🔗 View workflow run