fix(dcp): build the indexer query-split topology for a single shard

[koush]

VLLM_DCP_INDEXER_SHARDS=1 replicates the sparse-indexer K cache on every DCP rank, but the indexer topology was only built for 1 < shards < DCP, so one-shard runs fell back to _QUERY_SPLIT, which is singleton at TP == DCP. Every rank then scored every query row against the full indexer cache during prefill: DCP-wide redundant compute (~8x at DCP8, measurable as a long-context prefill throughput loss). With VLLM_DCP_QUERY_SPLIT=1 it was worse than redundant, since get_indexer_query_split_group(1) had no group to return and raised "No indexer query-split group matches the requested KV shard count".

_build_indexer_replica_group_ranks already yields the correct layout for shards=1 -- single-rank owner groups and one TP-wide query-split group -- so let it run. Each rank scores totalQ/TP rows against the full replicated cache, and the existing indices-only all-gather restores the shared top-k buffer. No merge is needed or attempted: dcp_world_size stays 1 for a replicated cache, so every candidate-merge path still early-returns.

Claude-Session: https://claude.ai/code/session_01J4BQPVrK6LdTUbXudyPkga

PLEASE FILL IN THE PR DESCRIPTION HERE ENSURING ALL CHECKLIST ITEMS (AT THE BOTTOM) HAVE BEEN CONSIDERED.

Purpose

Test Plan

Test Result

---

Essential Elements of an Effective PR Description Checklist

• The purpose of the PR, such as "Fix some issue (link existing issues this PR will resolve)".
• The test plan, such as providing test command.
• The test results, such as pasting the results comparison before and after, or e2e results
• (Optional) The necessary documentation update, such as updating supported_models.md and examples for a new model.

BEFORE SUBMITTING, PLEASE READ https://docs.vllm.ai/en/latest/contributing (anything written below this line will be removed by GitHub Actions)

Summary by CodeRabbit

• Bug Fixes
  • Improved distributed processing initialization for configurations using a single indexer shard with multiple decode-context parallel partitions.
  • Ensures these configurations now initialize correctly and operate as expected.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro Plus

Run ID: 58659ecb-a678-413b-8999-fa6fb5b488c9

📥 Commits

Reviewing files that changed from the base of the PR and between 3003860 and 52cece6.

📒 Files selected for processing (1)

• vllm/distributed/parallel_state.py

---

📝 Walkthrough

Walkthrough

The indexer replica-group initialization condition now supports indexer_shards == 1 when the decode-context parallel size is greater than one.

Changes

Indexer DCP initialization

Layer / File(s)	Summary
Expand indexer replica-group condition vllm/distributed/parallel_state.py	The setup condition includes single-shard indexer configurations with multi-way DCP. Zero and full DCP shard counts remain excluded.

Estimated code review effort: 1 (Trivial) | ~5 minutes

Possibly related PRs

• local-inference-lab/vllm#175: Updates related indexer DCP group initialization and topology handling.
• local-inference-lab/vllm#176: Modifies related DCP and query-split group initialization in parallel_state.py.

Suggested reviewers: voipmonitor, lukealonso, yewentao256

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title clearly and concisely describes the fix for building DCP indexer query-split topology with a single shard.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[voipmonitor]

Superseded by #245. The original functional commit is preserved unchanged; the replacement adds focused regression coverage for the indexer_shards=1 initialization boundary and documents the completed TP4/DCP4 topology and A-B-A E2E validation. A separate PR was required because GitHub rejected a direct maintainer push to this fork branch with HTTP 403 despite maintainer_can_modify=true.