search

vllm-project / vllm

A high-throughput and memory-efficient inference and serving engine for LLMs
https://docs.vllm.ai
Apache License 2.0
30.69k stars 4.66k forks source link

[Bug][0.5.4] Front-end server errors when overloaded with pending requests #7309

Open mgoin opened 3 months ago

mgoin commented 3 months ago

Your current environment

The output of python collect_env.py:

```text PyTorch version: 2.4.0+cu121 Is debug build: False CUDA used to build PyTorch: 12.1 ROCM used to build PyTorch: N/A OS: Ubuntu 22.04.4 LTS (x86_64) GCC version: (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0 Clang version: Could not collect CMake version: version 3.30.2 Libc version: glibc-2.35 Python version: 3.10.12 (main, Nov 20 2023, 15:14:05) [GCC 11.4.0] (64-bit runtime) Python platform: Linux-5.15.0-107-generic-x86_64-with-glibc2.35 Is CUDA available: True CUDA runtime version: 12.5.82 CUDA_MODULE_LOADING set to: LAZY GPU models and configuration: GPU 0: NVIDIA A100-SXM4-80GB GPU 1: NVIDIA A100-SXM4-80GB GPU 2: NVIDIA A100-SXM4-80GB GPU 3: NVIDIA A100-SXM4-80GB GPU 4: NVIDIA A100-SXM4-80GB GPU 5: NVIDIA A100-SXM4-80GB GPU 6: NVIDIA A100-SXM4-80GB GPU 7: NVIDIA A100-SXM4-80GB Nvidia driver version: 550.54.15 cuDNN version: Could not collect HIP runtime version: N/A MIOpen runtime version: N/A Is XNNPACK available: True CPU: Architecture: x86_64 CPU op-mode(s): 32-bit, 64-bit Address sizes: 46 bits physical, 57 bits virtual Byte Order: Little Endian CPU(s): 128 On-line CPU(s) list: 0-127 Vendor ID: GenuineIntel Model name: Intel(R) Xeon(R) Platinum 8462Y+ CPU family: 6 Model: 143 Thread(s) per core: 2 Core(s) per socket: 32 Socket(s): 2 Stepping: 8 BogoMIPS: 5600.00 Flags: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx pdpe1gb rdtscp lm constant_tsc art arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc cpuid aperfmperf tsc_known_freq pni pclmulqdq dtes64 monitor ds_cpl vmx smx est tm2 ssse3 sdbg fma cx16 xtpr pdcm pcid dca sse4_1 sse4_2 x2apic movbe popcnt tsc_deadline_timer aes xsave avx f16c rdrand lahf_lm abm 3dnowprefetch cpuid_fault epb cat_l3 cat_l2 cdp_l3 invpcid_single cdp_l2 ssbd mba ibrs ibpb stibp ibrs_enhanced tpr_shadow vnmi flexpriority ept vpid ept_ad fsgsbase tsc_adjust bmi1 avx2 smep bmi2 erms invpcid cqm rdt_a avx512f avx512dq rdseed adx smap avx512ifma clflushopt clwb intel_pt avx512cd sha_ni avx512bw avx512vl xsaveopt xsavec xgetbv1 xsaves cqm_llc cqm_occup_llc cqm_mbm_total cqm_mbm_local split_lock_detect avx_vnni avx512_bf16 wbnoinvd dtherm ida arat pln pts avx512vbmi umip pku ospke waitpkg avx512_vbmi2 gfni vaes vpclmulqdq avx512_vnni avx512_bitalg tme avx512_vpopcntdq la57 rdpid bus_lock_detect cldemote movdiri movdir64b enqcmd fsrm md_clear serialize tsxldtrk pconfig arch_lbr amx_bf16 avx512_fp16 amx_tile amx_int8 flush_l1d arch_capabilities Virtualization: VT-x L1d cache: 3 MiB (64 instances) L1i cache: 2 MiB (64 instances) L2 cache: 128 MiB (64 instances) L3 cache: 120 MiB (2 instances) NUMA node(s): 2 NUMA node0 CPU(s): 0,2,4,6,8,10,12,14,16,18,20,22,24,26,28,30,32,34,36,38,40,42,44,46,48,50,52,54,56,58,60,62,64,66,68,70,72,74,76,78,80,82,84,86,88,90,92,94,96,98,100,102,104,106,108,110,112,114,116,118,120,122,124,126 NUMA node1 CPU(s): 1,3,5,7,9,11,13,15,17,19,21,23,25,27,29,31,33,35,37,39,41,43,45,47,49,51,53,55,57,59,61,63,65,67,69,71,73,75,77,79,81,83,85,87,89,91,93,95,97,99,101,103,105,107,109,111,113,115,117,119,121,123,125,127 Vulnerability Gather data sampling: Not affected Vulnerability Itlb multihit: Not affected Vulnerability L1tf: Not affected Vulnerability Mds: Not affected Vulnerability Meltdown: Not affected Vulnerability Mmio stale data: Not affected Vulnerability Retbleed: Not affected Vulnerability Spec rstack overflow: Not affected Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl and seccomp Vulnerability Spectre v1: Mitigation; usercopy/swapgs barriers and __user pointer sanitization Vulnerability Spectre v2: Mitigation; Enhanced IBRS; IBPB conditional; RSB filling; PBRSB-eIBRS SW sequence; BHI BHI_DIS_S Vulnerability Srbds: Not affected Vulnerability Tsx async abort: Not affected Versions of relevant libraries: [pip3] numpy==1.26.4 [pip3] nvidia-nccl-cu12==2.20.5 [pip3] torch==2.4.0 [pip3] torchvision==0.19.0 [pip3] transformers==4.44.0 [pip3] triton==3.0.0 [conda] Could not collect ROCM Version: Could not collect Neuron SDK Version: N/A vLLM Version: 0.5.4 vLLM Build Flags: CUDA Archs: Not Set; ROCm: Disabled; Neuron: Disabled GPU Topology: GPU0 GPU1 GPU2 GPU3 GPU4 GPU5 GPU6 GPU7 NIC0 NIC1 NIC2 NIC3 NIC4 NIC5 NIC6 NIC7 CPU Affinity NUMA Affinity GPU NUMA ID GPU0 X NV12 NV12 NV12 NV12 NV12 NV12 NV12 PXB SYS SYS SYS SYS SYS SYS SYS 0,2,4,6,8,10 0 N/A GPU1 NV12 X NV12 NV12 NV12 NV12 NV12 NV12 SYS PXB SYS SYS SYS SYS SYS SYS 0,2,4,6,8,10 0 N/A GPU2 NV12 NV12 X NV12 NV12 NV12 NV12 NV12 SYS SYS PXB SYS SYS SYS SYS SYS 0,2,4,6,8,10 0 N/A GPU3 NV12 NV12 NV12 X NV12 NV12 NV12 NV12 SYS SYS SYS PXB SYS SYS SYS SYS 0,2,4,6,8,10 0 N/A GPU4 NV12 NV12 NV12 NV12 X NV12 NV12 NV12 SYS SYS SYS SYS PXB SYS SYS SYS 1,3,5,7,9,11 1 N/A GPU5 NV12 NV12 NV12 NV12 NV12 X NV12 NV12 SYS SYS SYS SYS SYS PXB SYS SYS 1,3,5,7,9,11 1 N/A GPU6 NV12 NV12 NV12 NV12 NV12 NV12 X NV12 SYS SYS SYS SYS SYS SYS PXB SYS 1,3,5,7,9,11 1 N/A GPU7 NV12 NV12 NV12 NV12 NV12 NV12 NV12 X SYS SYS SYS SYS SYS SYS SYS PXB 1,3,5,7,9,11 1 N/A NIC0 PXB SYS SYS SYS SYS SYS SYS SYS X SYS SYS SYS SYS SYS SYS SYS NIC1 SYS PXB SYS SYS SYS SYS SYS SYS SYS X SYS SYS SYS SYS SYS SYS NIC2 SYS SYS PXB SYS SYS SYS SYS SYS SYS SYS X SYS SYS SYS SYS SYS NIC3 SYS SYS SYS PXB SYS SYS SYS SYS SYS SYS SYS X SYS SYS SYS SYS NIC4 SYS SYS SYS SYS PXB SYS SYS SYS SYS SYS SYS SYS X SYS SYS SYS NIC5 SYS SYS SYS SYS SYS PXB SYS SYS SYS SYS SYS SYS SYS X SYS SYS NIC6 SYS SYS SYS SYS SYS SYS PXB SYS SYS SYS SYS SYS SYS SYS X SYS NIC7 SYS SYS SYS SYS SYS SYS SYS PXB SYS SYS SYS SYS SYS SYS SYS X Legend: X = Self SYS = Connection traversing PCIe as well as the SMP interconnect between NUMA nodes (e.g., QPI/UPI) NODE = Connection traversing PCIe as well as the interconnect between PCIe Host Bridges within a NUMA node PHB = Connection traversing PCIe as well as a PCIe Host Bridge (typically the CPU) PXB = Connection traversing multiple PCIe bridges (without traversing the PCIe Host Bridge) PIX = Connection traversing at most a single PCIe bridge NV# = Connection traversing a bonded set of # NVLinks NIC Legend: NIC0: mlx5_0 NIC1: mlx5_1 NIC2: mlx5_2 NIC3: mlx5_3 NIC4: mlx5_4 NIC5: mlx5_5 NIC6: mlx5_6 NIC7: mlx5_7 ```

🐛 Describe the bug

It seems the front-end server can easily get overloaded when there are many pending requests (>1000 seems to roughly be the threshold).

Individual requests over the threshold being failing with:

ERROR:    Exception in ASGI application
Traceback (most recent call last):
  File "/home/mgoin/venvs/vllm-rel/lib/python3.10/site-packages/starlette/responses.py", line 265, in __call__
    await wrap(partial(self.listen_for_disconnect, receive))
  File "/home/mgoin/venvs/vllm-rel/lib/python3.10/site-packages/starlette/responses.py", line 261, in wrap
    await func()
  File "/home/mgoin/venvs/vllm-rel/lib/python3.10/site-packages/starlette/responses.py", line 238, in listen_for_disconnect
    message = await receive()
  File "/home/mgoin/venvs/vllm-rel/lib/python3.10/site-packages/uvicorn/protocols/http/httptools_impl.py", line 553, in receive
    await self.message_event.wait()
  File "/usr/lib/python3.10/asyncio/locks.py", line 214, in wait
    await fut
asyncio.exceptions.CancelledError: Cancelled by cancel scope 7f6028dc8ee0

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "/home/mgoin/venvs/vllm-rel/lib/python3.10/site-packages/uvicorn/protocols/http/httptools_impl.py", line 399, in run_asgi
    result = await app(  # type: ignore[func-returns-value]
  File "/home/mgoin/venvs/vllm-rel/lib/python3.10/site-packages/uvicorn/middleware/proxy_headers.py", line 70, in __call__
    return await self.app(scope, receive, send)
  File "/home/mgoin/venvs/vllm-rel/lib/python3.10/site-packages/fastapi/applications.py", line 1054, in __call__
    await super().__call__(scope, receive, send)
  File "/home/mgoin/venvs/vllm-rel/lib/python3.10/site-packages/starlette/applications.py", line 123, in __call__
    await self.middleware_stack(scope, receive, send)
  File "/home/mgoin/venvs/vllm-rel/lib/python3.10/site-packages/starlette/middleware/errors.py", line 186, in __call__
    raise exc
  File "/home/mgoin/venvs/vllm-rel/lib/python3.10/site-packages/starlette/middleware/errors.py", line 164, in __call__
    await self.app(scope, receive, _send)
  File "/home/mgoin/venvs/vllm-rel/lib/python3.10/site-packages/starlette/middleware/cors.py", line 85, in __call__
    await self.app(scope, receive, send)
  File "/home/mgoin/venvs/vllm-rel/lib/python3.10/site-packages/starlette/middleware/exceptions.py", line 65, in __call__
    await wrap_app_handling_exceptions(self.app, conn)(scope, receive, send)
  File "/home/mgoin/venvs/vllm-rel/lib/python3.10/site-packages/starlette/_exception_handler.py", line 64, in wrapped_app
    raise exc
  File "/home/mgoin/venvs/vllm-rel/lib/python3.10/site-packages/starlette/_exception_handler.py", line 53, in wrapped_app
    await app(scope, receive, sender)
  File "/home/mgoin/venvs/vllm-rel/lib/python3.10/site-packages/starlette/routing.py", line 756, in __call__
    await self.middleware_stack(scope, receive, send)
  File "/home/mgoin/venvs/vllm-rel/lib/python3.10/site-packages/starlette/routing.py", line 776, in app
    await route.handle(scope, receive, send)
  File "/home/mgoin/venvs/vllm-rel/lib/python3.10/site-packages/starlette/routing.py", line 297, in handle
    await self.app(scope, receive, send)
  File "/home/mgoin/venvs/vllm-rel/lib/python3.10/site-packages/starlette/routing.py", line 77, in app
    await wrap_app_handling_exceptions(app, request)(scope, receive, send)
  File "/home/mgoin/venvs/vllm-rel/lib/python3.10/site-packages/starlette/_exception_handler.py", line 64, in wrapped_app
    raise exc
  File "/home/mgoin/venvs/vllm-rel/lib/python3.10/site-packages/starlette/_exception_handler.py", line 53, in wrapped_app
    await app(scope, receive, sender)
  File "/home/mgoin/venvs/vllm-rel/lib/python3.10/site-packages/starlette/routing.py", line 75, in app
    await response(scope, receive, send)
  File "/home/mgoin/venvs/vllm-rel/lib/python3.10/site-packages/starlette/responses.py", line 258, in __call__
    async with anyio.create_task_group() as task_group:
  File "/home/mgoin/venvs/vllm-rel/lib/python3.10/site-packages/anyio/_backends/_asyncio.py", line 680, in __aexit__
    raise BaseExceptionGroup(
exceptiongroup.ExceptionGroup: unhandled errors in a TaskGroup (1 sub-exception)

This quickly fills the server's output as it throw an exception for each request.

Steps to replicate

Server command:

vllm serve meta-llama/Meta-Llama-3.1-8B-Instruct --disable-log-requests

Benchmark command (needs more than 1000 pending prompts to trigger):

python benchmarks/benchmark_serving.py --model meta-llama/Meta-Llama-3.1-8B-Instruct --dataset-name random --port 8000 --num-prompts 2000

NOTE: The backend engine seems to continue running fine, it is just new requests throw exceptions in the front-end

  File "/home/mgoin/venvs/vllm-rel/lib/python3.10/site-packages/starlette/responses.py", line 258, in __call__
    async with anyio.create_task_group() as task_group:
  File "/home/mgoin/venvs/vllm-rel/lib/python3.10/site-packages/anyio/_backends/_asyncio.py", line 680, in __aexit__
    raise BaseExceptionGroup(
exceptiongroup.ExceptionGroup: unhandled errors in a TaskGroup (1 sub-exception)
INFO 08-08 16:36:33 metrics.py:406] Avg prompt throughput: 10724.5 tokens/s, Avg generation throughput: 651.8 tokens/s, Running: 77 reqs, Swapped: 0 reqs, Pending: 941 reqs, GPU KV cache usage: 19.3%, CPU KV cache usage: 0.0%.
INFO 08-08 16:36:38 metrics.py:406] Avg prompt throughput: 4585.6 tokens/s, Avg generation throughput: 480.5 tokens/s, Running: 96 reqs, Swapped: 0 reqs, Pending: 920 reqs, GPU KV cache usage: 24.3%, CPU KV cache usage: 0.0%.
INFO 08-08 16:36:43 metrics.py:406] Avg prompt throughput: 9370.5 tokens/s, Avg generation throughput: 1124.9 tokens/s, Running: 95 reqs, Swapped: 0 reqs, Pending: 875 reqs, GPU KV cache usage: 24.0%, CPU KV cache usage: 0.0%.
INFO 08-08 16:36:48 metrics.py:406] Avg prompt throughput: 8883.0 tokens/s, Avg generation throughput: 1006.5 tokens/s, Running: 90 reqs, Swapped: 0 reqs, Pending: 833 reqs, GPU KV cache usage: 22.7%, CPU KV cache usage: 0.0%.
INFO 08-08 16:36:53 metrics.py:406] Avg prompt throughput: 9190.6 tokens/s, Avg generation throughput: 1026.0 tokens/s, Running: 92 reqs, Swapped: 0 reqs, Pending: 790 reqs, GPU KV cache usage: 23.3%, CPU KV cache usage: 0.0%.

Running with --disable-frontend-multiprocessing or downgrading to v0.5.3 will resolve the issue.

github-actions[bot] commented 2 weeks ago

This issue has been automatically marked as stale because it has not had any activity within 90 days. It will be automatically closed if no further activity occurs within 30 days. Leave a comment if you feel this issue should remain open. Thank you!