[PD] bug fix: Update status if nixl receiver send a a dummy req. #6720

thesues · 2025-05-28T23:29:21Z

For PD, if Decode's tp-size is less than Prefill's tp-size. Prefill will throw this exception.

Exception in thread Thread-5 (bootstrap_thread):
Traceback (most recent call last):
  File "/usr/lib/python3.10/threading.py", line 1016, in _bootstrap_inner
Exception in thread Thread-5 (bootstrap_thread):
Traceback (most recent call last):
  File "/usr/lib/python3.10/threading.py", line 1016, in _bootstrap_inner
Exception in thread Thread-5 (bootstrap_thread):
    self.run()
  File "/usr/lib/python3.10/threading.py", line 953, in run
Traceback (most recent call last):
  File "/usr/lib/python3.10/threading.py", line 1016, in _bootstrap_inner
    self.run()
  File "/usr/lib/python3.10/threading.py", line 953, in run
    self._target(*self._args, **self._kwargs)
  File "/sgl-workspace/sglang_deepep/python/sglang/srt/disaggregation/nixl/conn.py", line 351, in bootstrap_thread
    self._target(*self._args, **self._kwargs)
  File "/sgl-workspace/sglang_deepep/python/sglang/srt/disaggregation/nixl/conn.py", line 351, in bootstrap_thread
    self.run()
    required_dst_info_num = int(waiting_req_bytes[10].decode("ascii"))
    required_dst_info_num = int(waiting_req_bytes[10].decode("ascii"))
IndexError: list index out of range
  File "/usr/lib/python3.10/threading.py", line 953, in run
IndexError: list index out of range
    self._target(*self._args, **self._kwargs)
  File "/sgl-workspace/sglang_deepep/python/sglang/srt/disaggregation/nixl/conn.py", line 351, in bootstrap_thread
    required_dst_info_num = int(waiting_req_bytes[10].decode("ascii"))
IndexError: list index out of range

The nixl PREFILL manager should detect the dummy msg by counting waiting_req_bytes, and update status to WaitingForInput. My test is OK for these commands:

PREFILL/TP4

python -m sglang.launch_server --model-path /data/models/DeepSeek-V2-Lite/ --port 40001 --host 127.0.0.1  --tp-size 4  --trust-remote-code --disaggregation-mode prefill --page-size 32 --base-gpu-id 0 --disaggregation-transfer-backend nixl --disable-radix-cache

DECODE/TP2

python -m sglang.launch_server --model-path /data/models/DeepSeek-V2-Lite/ --port 40100 --host 127.0.0.1  --tp-size 2--trust-remote-code --disaggregation-mode decode --page-size 32 --base-gpu-id 4 --disaggregation-transfer-backend nixl --disable-radix-cache

gemini-code-assist

Hello @thesues, I'm Gemini Code Assist¹! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed!

Summary of Changes

Hello! Gemini here, providing a summary of this pull request to help everyone get up to speed quickly.

This pull request aims to fix a specific IndexError: list index out of range that occurs in the Nixl disaggregation mode when the Decode process is configured with a smaller tensor parallel (TP) size than the Prefill process. The error happens in the bootstrap_thread function when the Prefill side receives a short, seemingly 'dummy' message from the Decode side and attempts to parse it as a full request.

The core change is to add logic to detect these short/dummy messages based on the message length (len(waiting_req_bytes) == 1) and handle them gracefully by updating the status for the corresponding room to WaitingForInput instead of attempting to parse the message further, which prevents the IndexError.

Highlights

Bug Fix: Addresses an IndexError in the Nixl disaggregation backend when Decode TP size is less than Prefill TP size.
Dummy Message Handling: Introduces specific logic in the bootstrap_thread to detect and handle short 'dummy' messages sent between disaggregated components.
Status Update: Updates the status of a connection room to WaitingForInput upon receiving a detected dummy message, allowing the process to continue correctly.

Changelog

python/sglang/srt/disaggregation/nixl/conn.py
- Modified the bootstrap_thread function to correctly handle short messages.
- Added a check (if len(waiting_req_bytes) == 1:) to identify dummy messages.
- For dummy messages, the status for the corresponding room is updated to KVPoll.WaitingForInput.
- The parsing of required_dst_info_num is now performed only after confirming the message is not a dummy message, preventing the IndexError.

Using Gemini Code Assist

The full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips.

Invoking Gemini

You can request assistance from Gemini at any point in your pull request via creating an issue comment (i.e. comment on the pull request page) using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands.

Feature	Command	Description
Code Review	`/gemini review`	Performs a code review for the current pull request in its current state.
Pull Request Summary	`/gemini summary`	Provides a summary of the current pull request in its current state.
Comment	@gemini-code-assist	Responds in comments when explicitly tagged, both in issue comments and review comments.
Help	`/gemini help`	Displays a list of available commands.

Customization

To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a .gemini/ folder in the base of the repository. Detailed instructions can be found here.

Limitations & Feedback

Gemini Code Assist is currently in preview and may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments to provide feedback.

You can also get AI-powered code generation, chat, as well as code reviews directly in the IDE at no cost with the Gemini Code Assist IDE Extension.

A byte list too short,
Index error, a bad sort.
Check length, then parse.

Review the Privacy Notices, Generative AI Prohibited Use Policy, Terms of Service, and learn how to configure Gemini Code Assist in GitHub here. Gemini can make mistakes, so double check it and use code with caution. ↩

gemini-code-assist

Code Review

This pull request effectively addresses the IndexError in the bootstrap_thread by adding a check for dummy messages based on the length of waiting_req_bytes. The logic to handle these dummy messages by updating transfer_infos and setting the status to WaitingForInput appears correct and aligns with the expected behavior for such messages. The fix is well-targeted to the issue described.

I have one suggestion for a minor cleanup to improve code clarity.

Summary of Findings

Potential IndexError: The code accesses waiting_req_bytes[10] without ensuring that waiting_req_bytes has at least 11 elements, which could lead to an IndexError.
Missing Input Validation: The code converts user input to an integer without proper validation, which could lead to a ValueError and program termination.

Merge Readiness

The pull request addresses the bug effectively, but there are a couple of suggestions for improvement. I would recommend addressing these before merging. However, as an AI assistant, I am not authorized to approve pull requests. Please ensure further review and approval from the repository maintainers before merging.

python/sglang/srt/disaggregation/nixl/conn.py

thesues · 2025-05-29T00:32:09Z

@trevor-m I updated the PR. This fix is very simple as you said. test is OK

set kv_indices b"" if dummy
update is_dummy function

trevor-m · 2025-05-29T00:36:09Z

@trevor-m I updated the PR. This fix is very simple as you said. test is OK

set kv_indices b"" if dummy

update is_dummy function

Thank you! It looks good.

I found another bug with dummy req, you may also need this one line change: https://github.com/sgl-project/sglang/pull/6717/files#diff-e65bb073f9d45fe11752f91234d1fd319d542a4008b90123fc0408acf40676d7R299

zhyncs · 2025-05-29T00:41:52Z

@thesues pre-commit run --all-files

thesues · 2025-05-29T03:11:43Z

@thesues pre-commit run --all-files

done.

thesues · 2025-05-29T03:12:54Z

@trevor-m I updated the PR. This fix is very simple as you said. test is OK

set kv_indices b"" if dummy

update is_dummy function

Thank you! It looks good.

I found another bug with dummy req, you may also need this one line change: https://github.com/sgl-project/sglang/pull/6717/files#diff-e65bb073f9d45fe11752f91234d1fd319d542a4008b90123fc0408acf40676d7R299

sure, it's updated

…-project#6720)

Merge branch 'sgl_20250610_sync_tag047 of git@code.alipay.com:Theta/SGLang.git into main https://code.alipay.com/Theta/SGLang/pull_requests/52 Reviewed-by: 剑川 <jianchuan.gys@antgroup.com> * [Bugfix] Fix slice operation when chunk size mismatch (sgl-project#6697) * [Bugfix] Fix ChatCompletion endpoint of mini_lb when stream is set (sgl-project#6703) * [CI] Fix setup of disaggregation with different tp (sgl-project#6706) * [PD] Remove Unnecessary Exception Handling for FastQueue.get() (sgl-project#6712) * Fuse routed_scaling_factor in DeepSeek (sgl-project#6710) * Overlap two kernels in DeepSeek with communication (sgl-project#6711) * Minor refactor two-batch overlap (sgl-project#6682) * Speed up when having padding tokens two-batch overlap (sgl-project#6668) * [Feature] Support Flashinfer fp8 blockwise GEMM kernel on Blackwell (sgl-project#6479) * Fix LoRA bench (sgl-project#6719) * temp * Fix PP for Qwen3 MoE (sgl-project#6709) * [feat] triton kernel for get_last_loc (sgl-project#6676) * [fix] more mem for draft_extend cuda_graph (sgl-project#6726) * [PD] bug fix: Update status if nixl receiver send a a dummy req. (sgl-project#6720) * Tune memory arguments on B200 (sgl-project#6718) * Add DeepSeek-R1-0528 function call chat template (sgl-project#6725) * refactor(tool call): Fix BaseFormatDetector tool_index issue and refactor `parse_streaming_increment` (sgl-project#6715) * Add draft extend CUDA graph for Triton backend (sgl-project#6705) * refactor apply_w8a8_block_fp8_linear in fp (sgl-project#6545) * [PD] Support completion endpoint (sgl-project#6729) * PD Rust LB (PO2) (sgl-project#6437) * Super tiny enable sole usage of expert distribution metrics and update doc (sgl-project#6680) * Support picking variants of EPLB algorithms (sgl-project#6728) * Support tuning DeepEP configs (sgl-project#6742) * [test] add ut and bm for get_last_loc (sgl-project#6746) * Fix mem_fraction_static for AMD CI (sgl-project#6748) * [fix][RL] Fix DeepSeekV3ForCausalLM.post_load_weights for multiple update weight (sgl-project#6265) * Improve EPLB logical to physical dispatch map (sgl-project#6727) * Update DeepSeek-R1-0528 function call chat template (sgl-project#6765) * [PD] Optimize time out logic and add env var doc for mooncake (sgl-project#6761) * Fix aiohttp 'Chunk too big' in bench_serving (sgl-project#6737) * Support sliding window in triton backend (sgl-project#6509) * Fix shared experts fusion error (sgl-project#6289) * Fix one bug in the grouped-gemm triton kernel (sgl-project#6772) * update llama4 chat template and pythonic parser (sgl-project#6679) * feat(tool call): Enhance Llama32Detector for improved JSON parsing in non-stream (sgl-project#6784) * Support token-level quantization for EP MoE (sgl-project#6782) * Temporarily lower mmlu threshold for triton sliding window backend (sgl-project#6785) * ci: relax test_function_call_required (sgl-project#6786) * Add intel_amx backend for Radix Attention for CPU (sgl-project#6408) * Fix incorrect LoRA weight loading for fused gate_up_proj (sgl-project#6734) * fix(PD-disaggregation): Can not get local ip (sgl-project#6792) * [FIX] mmmu bench serving result display error (sgl-project#6525) (sgl-project#6791) * Bump torch to 2.7.0 (sgl-project#6788) * chore: bump sgl-kernel v0.1.5 (sgl-project#6794) * Improve profiler and integrate profiler in bench_one_batch_server (sgl-project#6787) * chore: upgrade sgl-kernel v0.1.5 (sgl-project#6795) * [Minor] Always append newline after image token when parsing chat message (sgl-project#6797) * Update CI tests for Llama4 models (sgl-project#6421) * [Feat] Enable PDL automatically on Hopper architecture (sgl-project#5981) * chore: update blackwell docker (sgl-project#6800) * misc: cache is_hopper_arch (sgl-project#6799) * Remove contiguous before Flashinfer groupwise fp8 gemm (sgl-project#6804) * Correctly abort the failed grammar requests & Improve the handling of abort (sgl-project#6803) * [EP] Add cuda kernel for moe_ep_pre_reorder (sgl-project#6699) * Add draft extend CUDA graph for flashinfer backend (sgl-project#6805) * Refactor CustomOp to avoid confusing bugs (sgl-project#5382) * Tiny log prefill time (sgl-project#6780) * Tiny fix EPLB assertion about rebalancing period and recorder window size (sgl-project#6813) * Add simple utility to dump tensors for debugging (sgl-project#6815) * Fix profiles do not have consistent names (sgl-project#6811) * Speed up rebalancing when using non-static dispatch algorithms (sgl-project#6812) * [1/2] Add Kernel support for Cutlass based Fused FP4 MoE (sgl-project#6093) * [Router] Fix k8s Service Discovery (sgl-project#6766) * Add CPU optimized kernels for topk and rope fusions (sgl-project#6456) * fix new_page_count_next_decode (sgl-project#6671) * Fix wrong weight reference in dynamic EPLB (sgl-project#6818) * Minor add metrics to expert location updater (sgl-project#6816) * [Refactor] Rename `n_share_experts_fusion` as `num_fused_shared_experts` (sgl-project#6735) * [FEAT] Add transformers backend support (sgl-project#5929) * [fix] recover auto-dispatch for rmsnorm and rope (sgl-project#6745) * fix ep_moe_reorder kernel bugs (sgl-project#6858) * [Refactor] Multimodal data processing for VLM (sgl-project#6659) * Decoder-only Scoring API (sgl-project#6460) * feat: add dp-rank to KV events (sgl-project#6852) * Set `num_fused_shared_experts` as `num_shared_experts` when shared_experts fusion is not disabled (sgl-project#6736) * Fix one missing arg in DeepEP (sgl-project#6878) * Support LoRA in TestOpenAIVisionServer and fix fused kv_proj loading bug. (sgl-project#6861) * support 1 shot allreduce in 1-node and 2-node using mscclpp (sgl-project#6277) * Fix Qwen3MoE missing token padding optimization (sgl-project#6820) * Tiny update error hints (sgl-project#6846) * Support layerwise rebalancing experts (sgl-project#6851) * Tiny allow profiler API to auto create directory (sgl-project#6865) * Support Blackwell DeepEP docker images (sgl-project#6868) * [EP] Add cuda kernel for moe_ep_post_reorder (sgl-project#6837) * [theta]merge 0605 * oai: fix openAI client error with single request via batch api (sgl-project#6170) * [PD] Fix potential perf spike caused by tracker gc and optimize doc (sgl-project#6764) * Use deepgemm instead of triton for fused_qkv_a_proj_with_mqa (sgl-project#6890) * [CUTLASS-FP4-MOE] Introduce CutlassMoEParams class for easy initialization of Cutlass Grouped Gems Metadata (sgl-project#6887) * bugfix(OAI): Fix image_data processing for jinja chat templates (sgl-project#6877) * [CPU] enable CI for PRs, add Dockerfile and auto build task (sgl-project#6458) * AITER backend extension and workload optimizations (sgl-project#6838) * [theta]merge * [theta]merge * [Feature] Support Flashinfer fmha on Blackwell (sgl-project#6930) * Fix a bug in abort & Improve docstrings for abort (sgl-project#6931) * Tiny support customize DeepEP max dispatch tokens per rank (sgl-project#6934) * Sync the changes on cuda graph runners (sgl-project#6932) * [PD] Optimize transfer queue forward logic for dummy rank (sgl-project#6922) * [Refactor] image data process in bench_serving (sgl-project#6879) * [fix] logical_to_all_physical_map index 256 is out of bounds in EP parallel. (sgl-project#6767) * Add triton fused moe kernel config for E=257 on B200 (sgl-project#6939) * [sgl-kernel] update deepgemm (sgl-project#6942) * chore: bump sgl-kernel v0.1.6 (sgl-project#6943) * Minor compile fused topk (sgl-project#6944) * [Bugfix] pipeline parallelism and Eagle Qwen2 (sgl-project#6910) * Tiny re-introduce profile id logging (sgl-project#6912) * Add triton version as a fused_moe_triton config search key to avoid performace decrease in different Triton version (sgl-project#5955) * reduce torch.zeros overhead in moe align block size kernel (sgl-project#6369) * chore: upgrade sgl-kernel v0.1.6 (sgl-project#6945) * add fbgemm moe grouped gemm kernel benchmark (sgl-project#6924) * [Docker] Add docker file for SGL Router (sgl-project#6915) * Disabling mixed chunked prefill when eagle is enabled (sgl-project#6874) * Add canary for EPLB rebalancing (sgl-project#6895) * Refactor global_server_args_dict (sgl-project#6866) * Fuse routed scaling factor in topk_reduce kernel (sgl-project#6220) * Update server timeout time in AMD CI. (sgl-project#6953) * [misc] add is_cpu() (sgl-project#6950) * Add H20 fused MoE kernel tuning configs for DeepSeek-R1/V3 (sgl-project#6885) * Add a CUDA kernel for fusing mapping and weighted sum for MoE. (sgl-project#6916) * chore: bump sgl-kernel v0.1.6.post1 (sgl-project#6955) * chore: upgrade sgl-kernel v0.1.6.post1 (sgl-project#6957) * [DeepseekR1-FP4] Add Support for nvidia/DeepSeekR1-FP4 model (sgl-project#6853) * Revert "Fuse routed scaling factor in topk_reduce kernel (sgl-project#6220)" (sgl-project#6968) * [AMD] Add more tests to per-commit-amd (sgl-project#6926) * chore: bump sgl-kernel v0.1.7 (sgl-project#6963) * Slightly improve the sampler to skip unnecessary steps (sgl-project#6956) * rebase h20 fused_moe config (sgl-project#6966) * Fix CI and triton moe Configs (sgl-project#6974) * Remove unnecessary kernels of num_token_non_padded (sgl-project#6965) * Extend cuda graph capture bs for B200 (sgl-project#6937) * Fuse routed scaling factor in deepseek (sgl-project#6970) * Sync cuda graph runners (sgl-project#6976) * Fix draft extend ut stability with flush cache (sgl-project#6979) * Fix triton sliding window test case (sgl-project#6981) * Fix expert distribution dumping causes OOM (sgl-project#6967) * Minor remove one kernel for DeepSeek (sgl-project#6977) * [perf][sgl-kernel] extend cutlass_mla_decode to support num_head < 128 (sgl-project#6929) * Enable more unit tests for AMD CI. (sgl-project#6983) * Use torch.compile to fuse flash attention decode metadata preparation (sgl-project#6973) * Eliminate stream sync to speed up LoRA batch init (sgl-project#6960) * support qwen3 emebedding (sgl-project#6990) * Fix torch profiler bugs for bench_offline_throughput.py (sgl-project#6557) * chore: upgrade flashinfer v0.2.6.post1 jit (sgl-project#6958) * cleanup tmp dir (sgl-project#7007) * chore: update pr test xeon (sgl-project#7008) * Fix cutlass MLA gets almost zero accuracy (sgl-project#6998) * Update amd nightly models CI. (sgl-project#6992) * feat: add direct routing strategy to DP worker (sgl-project#6884) * Fallback to lower triton version for unfound fused moe configs (sgl-project#7013) * Fix torchvision version for Blackwell (sgl-project#7015) * Simplify prepare_extend_after_decode (sgl-project#6987) * Migrate to assertEqual (sgl-project#6741) * Fix torch version in blackwell dockerfile (sgl-project#7017) * chore: update pr test xeon (sgl-project#7018) * Update default settings for blackwell (sgl-project#7023) * Support both approximate and exact expert distribution collection (sgl-project#6964) * Add decode req pool (sgl-project#6980) * [theta]merge 0610 * [theta]merge 0610 * [CI] Add CI workflow for sgl-router docker build (sgl-project#7027) * Fix fused_moe triton configs (sgl-project#7029) * CPU: map changes from developing branch in sgl-kernel (sgl-project#6833) * chore: bump v0.4.7 (sgl-project#7038) * Update README.md (sgl-project#7040)

thesues requested review from hnyls2002 and ByronHsu as code owners May 28, 2025 23:29

gemini-code-assist bot reviewed May 28, 2025

View reviewed changes

gemini-code-assist bot suggested changes May 28, 2025

View reviewed changes

python/sglang/srt/disaggregation/nixl/conn.py Show resolved Hide resolved

python/sglang/srt/disaggregation/nixl/conn.py Outdated Show resolved Hide resolved

trevor-m suggested changes May 28, 2025

View reviewed changes

python/sglang/srt/disaggregation/nixl/conn.py Outdated Show resolved Hide resolved

thesues force-pushed the fix-dummy-request branch 2 times, most recently from 37dc2b5 to 1de4336 Compare May 29, 2025 00:29

trevor-m approved these changes May 29, 2025

View reviewed changes

[PD] fix dummy nixl msg

9d3b906

thesues force-pushed the fix-dummy-request branch from 3e548cb to 9d3b906 Compare May 29, 2025 03:08

Merge branch 'main' into fix-dummy-request

8062923

trevor-m approved these changes May 29, 2025

View reviewed changes

Merge branch 'main' into fix-dummy-request

e513cf9

zhyncs approved these changes May 29, 2025

View reviewed changes

zhyncs merged commit c459536 into sgl-project:main May 29, 2025
11 of 37 checks passed

trevor-m mentioned this pull request May 29, 2025

[PD] NIXL: Register kv args in advance and cleanup finished requests #6717

Merged

6 tasks

thesues added a commit to bytedance-iaas/sglang that referenced this pull request May 29, 2025

port upstream sgl-project#6720

910487a

Layssy pushed a commit to Layssy/sglang-iaas that referenced this pull request Jun 9, 2025

[PD] bug fix: Update status if nixl receiver send a a dummy req. (sgl…

213f424

…-project#6720)

xwu-intel pushed a commit to xwu-intel/sglang that referenced this pull request Jun 17, 2025

[PD] bug fix: Update status if nixl receiver send a a dummy req. (sgl…

e91e1b6

…-project#6720)

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

[PD] bug fix: Update status if nixl receiver send a a dummy req. #6720

[PD] bug fix: Update status if nixl receiver send a a dummy req. #6720

Uh oh!

thesues commented May 28, 2025 •

edited

Loading

Uh oh!

gemini-code-assist bot left a comment

Uh oh!

gemini-code-assist bot left a comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

thesues commented May 29, 2025

Uh oh!

trevor-m commented May 29, 2025

Uh oh!

zhyncs commented May 29, 2025

Uh oh!

thesues commented May 29, 2025

Uh oh!

thesues commented May 29, 2025

Uh oh!

Uh oh!

Uh oh!

[PD] bug fix: Update status if nixl receiver send a a dummy req. #6720

[PD] bug fix: Update status if nixl receiver send a a dummy req. #6720

Uh oh!

Conversation

thesues commented May 28, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

gemini-code-assist bot left a comment

Choose a reason for hiding this comment

Summary of Changes

Highlights

Changelog

Footnotes

Uh oh!

gemini-code-assist bot left a comment

Choose a reason for hiding this comment

Code Review

Summary of Findings

Merge Readiness

Uh oh!

Uh oh!

Uh oh!

Uh oh!

thesues commented May 29, 2025

Uh oh!

trevor-m commented May 29, 2025

Uh oh!

zhyncs commented May 29, 2025

Uh oh!

thesues commented May 29, 2025

Uh oh!

thesues commented May 29, 2025

Uh oh!

Uh oh!

Uh oh!

thesues commented May 28, 2025 •

edited

Loading