vllm - 💡(How to fix) Fix [Bug]: Engine crashes with AssertionError when prompt exceeds auto-fitted max_model_len (admission check missing) [1 comments, 2 participants]

vllm2026-04-19 18:04:10

ON THIS PAGE

Recommended Tools

×6

Utilities matched from this issue’s tags and category — try them while you read without losing context.

GitHub issue graph ai analysis

Paste a GitHub issue URL. We fetch that issue, discover linked issues from bodies/comments/timeline, collect linked pull requests, and produce a structured English report.

The report is written in English Markdown for sharing and archival.

GitHub issue URL

Helpful · Quick feedback

GitHub stats

vllm-project/vllm#40302•Fetched 2026-04-20 11:59:26

View on GitHub

Comments

Participants

Timeline

Reactions

Author

key4ng

Participants

key4ng

sandyhu533

Timeline (top)

commented ×1labeled ×1mentioned ×1subscribed ×1

Error Message

4000 tokens: 200 - OK 8000 tokens: 200 - OK 12000 tokens: 200 - OK 15000 tokens: 200 - OK 15500 tokens: 500 - EngineCore encountered an issue. See stack trace (above) for the root caus 16000 tokens: (connection refused — engine is dead, pod restarting)

Code Example

Docker image: vllm/vllm-openai:gemma4
Collecting environment information...
==============================
        System Info
==============================
OS                           : Ubuntu 22.04.5 LTS (x86_64)
GCC version                  : (Ubuntu 11.4.0-1ubuntu1~22.04.3) 11.4.0
Clang version                : Could not collect
CMake version                : Could not collect
Libc version                 : glibc-2.35

==============================
       PyTorch Info
==============================
PyTorch version              : 2.10.0+cu129
Is debug build               : False
CUDA used to build PyTorch   : 12.9
ROCM used to build PyTorch   : N/A
XPU used to build PyTorch    : N/A

==============================
      Python Environment
==============================
Python version               : 3.12.13 (main, Mar  4 2026, 09:23:07) [GCC 11.4.0] (64-bit runtime)
Python platform              : Linux-5.15.0-1074-oracle-x86_64-with-glibc2.35

==============================
       CUDA / GPU Info
==============================
Is CUDA available            : True
CUDA runtime version         : 12.9.86
CUDA_MODULE_LOADING set to   :
GPU models and configuration : GPU 0: NVIDIA H100 80GB HBM3
Nvidia driver version        : 580.126.09
cuDNN version                : Could not collect
HIP runtime version          : N/A
MIOpen runtime version       : N/A
Is XNNPACK available         : True

==============================
          CPU Info
==============================
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):                               224
On-line CPU(s) list:                  0-223
Vendor ID:                            GenuineIntel
Model name:                           Intel(R) Xeon(R) Platinum 8480+
CPU family:                           6
Model:                                143
Thread(s) per core:                   2
Core(s) per socket:                   56
Socket(s):                            2
Stepping:                             8
CPU max MHz:                          3800.0000
CPU min MHz:                          800.0000
BogoMIPS:                             4000.00
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 intel_ppin 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 hwp hwp_act_window hwp_epp hwp_pkg_req 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:                            5.3 MiB (112 instances)
L1i cache:                            3.5 MiB (112 instances)
L2 cache:                             224 MiB (112 instances)
L3 cache:                             210 MiB (2 instances)
NUMA node(s):                         2
NUMA node0 CPU(s):                    0-55,112-167
NUMA node1 CPU(s):                    56-111,168-223
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 Reg file data sampling: Not affected
Vulnerability Retbleed:               Not affected
Vulnerability Spec rstack overflow:   Not affected
Vulnerability Spec store bypass:      Vulnerable
Vulnerability Spectre v1:             Vulnerable: __user pointer sanitization and usercopy barriers only; no swapgs barriers
Vulnerability Spectre v2:             Vulnerable; IBPB: disabled; STIBP: disabled; PBRSB-eIBRS: Vulnerable; BHI: Vulnerable
Vulnerability Srbds:                  Not affected
Vulnerability Tsx async abort:        Not affected

==============================
Versions of relevant libraries
==============================
[pip3] flashinfer-python==0.6.6
[pip3] numpy==2.2.6
[pip3] nvidia-cublas-cu12==12.9.1.4
[pip3] nvidia-cuda-cupti-cu12==12.9.79
[pip3] nvidia-cuda-nvrtc-cu12==12.9.86
[pip3] nvidia-cuda-runtime-cu12==12.9.79
[pip3] nvidia-cudnn-cu12==9.10.2.21
[pip3] nvidia-cudnn-frontend==1.18.0
[pip3] nvidia-cufft-cu12==11.4.1.4
[pip3] nvidia-cufile-cu12==1.14.1.1
[pip3] nvidia-curand-cu12==10.3.10.19
[pip3] nvidia-cusolver-cu12==11.7.5.82
[pip3] nvidia-cusparse-cu12==12.5.10.65
[pip3] nvidia-cusparselt-cu12==0.7.1
[pip3] nvidia-cutlass-dsl==4.4.2
[pip3] nvidia-cutlass-dsl-libs-base==4.4.2
[pip3] nvidia-ml-py==13.595.45
[pip3] nvidia-nccl-cu12==2.27.5
[pip3] nvidia-nvjitlink-cu12==12.9.86
[pip3] nvidia-nvshmem-cu12==3.4.5
[pip3] nvidia-nvtx-cu12==12.9.79
[pip3] pyzmq==27.1.0
[pip3] torch==2.10.0+cu129
[pip3] torch_c_dlpack_ext==0.1.5
[pip3] torchaudio==2.10.0+cu129
[pip3] torchvision==0.25.0+cu129
[pip3] transformers==5.5.0
[pip3] triton==3.6.0
[conda] Could not collect

==============================
         vLLM Info
==============================
ROCM Version                 : Could not collect
vLLM Version                 : 0.19.1.dev6+g6d4a8e6d2 (git sha: 6d4a8e6d2)
vLLM Build Flags:
  CUDA Archs: 7.0 7.5 8.0 8.9 9.0 10.0 12.0; ROCm: Disabled; XPU: Disabled
GPU Topology:
  	GPU0	NIC0	NIC1	NIC2	NIC3	NIC4	NIC5	NIC6	NIC7	NIC8	NIC9	NIC10	NIC11	NIC12	NIC13	NIC14	NIC15	NIC16	NIC17	CPU Affinity	NUMA Affinity	GPU NUMA ID
GPU0	 X 	NODE	NODE	NODE	NODE	NODE	NODE	NODE	PXB	PXB	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	0-55,112-167	0		N/A
NIC0	NODE	 X 	PIX	NODE	NODE	NODE	NODE	NODE	NODE	NODE	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS
NIC1	NODE	PIX	 X 	NODE	NODE	NODE	NODE	NODE	NODE	NODE	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS
NIC2	NODE	NODE	NODE	 X 	NODE	NODE	NODE	NODE	NODE	NODE	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS
NIC3	NODE	NODE	NODE	NODE	 X 	PIX	NODE	NODE	NODE	NODE	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS
NIC4	NODE	NODE	NODE	NODE	PIX	 X 	NODE	NODE	NODE	NODE	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS
NIC5	NODE	NODE	NODE	NODE	NODE	NODE	 X 	PIX	NODE	NODE	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS
NIC6	NODE	NODE	NODE	NODE	NODE	NODE	PIX	 X 	NODE	NODE	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS
NIC7	PXB	NODE	NODE	NODE	NODE	NODE	NODE	NODE	 X 	PIX	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS
NIC8	PXB	NODE	NODE	NODE	NODE	NODE	NODE	NODE	PIX	 X 	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS
NIC9	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	 X 	PIX	NODE	NODE	NODE	NODE	NODE	NODE	NODE
NIC10	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	PIX	 X 	NODE	NODE	NODE	NODE	NODE	NODE	NODE
NIC11	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	NODE	NODE	 X 	NODE	NODE	NODE	NODE	NODE	NODE
NIC12	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	NODE	NODE	NODE	 X 	PIX	NODE	NODE	NODE	NODE
NIC13	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	NODE	NODE	NODE	PIX	 X 	NODE	NODE	NODE	NODE
NIC14	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	NODE	NODE	NODE	NODE	NODE	 X 	PIX	NODE	NODE
NIC15	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	NODE	NODE	NODE	NODE	NODE	PIX	 X 	NODE	NODE
NIC16	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	NODE	NODE	NODE	NODE	NODE	NODE	NODE	 X 	PIX
NIC17	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	NODE	NODE	NODE	NODE	NODE	NODE	NODE	PIX	 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
  NIC8: mlx5_8
  NIC9: mlx5_9
  NIC10: mlx5_10
  NIC11: mlx5_11
  NIC12: mlx5_12
  NIC13: mlx5_13
  NIC14: mlx5_14
  NIC15: mlx5_15
  NIC16: mlx5_16
  NIC17: mlx5_17

==============================
     Environment Variables
==============================
NVIDIA_VISIBLE_DEVICES=GPU-960f71eb-1b3d-f6d2-bfcb-1a580f51b809
VLLM_RPC_TIMEOUT=30000
VLLM_ENGINE_ITERATION_TIMEOUT_S=120
NVIDIA_REQUIRE_CUDA=cuda>=12.9 brand=unknown,driver>=535,driver<536 brand=grid,driver>=535,driver<536 brand=tesla,driver>=535,driver<536 brand=nvidia,driver>=535,driver<536 brand=quadro,driver>=535,driver<536 brand=quadrortx,driver>=535,driver<536 brand=nvidiartx,driver>=535,driver<536 brand=vapps,driver>=535,driver<536 brand=vpc,driver>=535,driver<536 brand=vcs,driver>=535,driver<536 brand=vws,driver>=535,driver<536 brand=cloudgaming,driver>=535,driver<536 brand=unknown,driver>=550,driver<551 brand=grid,driver>=550,driver<551 brand=tesla,driver>=550,driver<551 brand=nvidia,driver>=550,driver<551 brand=quadro,driver>=550,driver<551 brand=quadrortx,driver>=550,driver<551 brand=nvidiartx,driver>=550,driver<551 brand=vapps,driver>=550,driver<551 brand=vpc,driver>=550,driver<551 brand=vcs,driver>=550,driver<551 brand=vws,driver>=550,driver<551 brand=cloudgaming,driver>=550,driver<551 brand=unknown,driver>=560,driver<561 brand=grid,driver>=560,driver<561 brand=tesla,driver>=560,driver<561 brand=nvidia,driver>=560,driver<561 brand=quadro,driver>=560,driver<561 brand=quadrortx,driver>=560,driver<561 brand=nvidiartx,driver>=560,driver<561 brand=vapps,driver>=560,driver<561 brand=vpc,driver>=560,driver<561 brand=vcs,driver>=560,driver<561 brand=vws,driver>=560,driver<561 brand=cloudgaming,driver>=560,driver<561 brand=unknown,driver>=565,driver<566 brand=grid,driver>=565,driver<566 brand=tesla,driver>=565,driver<566 brand=nvidia,driver>=565,driver<566 brand=quadro,driver>=565,driver<566 brand=quadrortx,driver>=565,driver<566 brand=nvidiartx,driver>=565,driver<566 brand=vapps,driver>=565,driver<566 brand=vpc,driver>=565,driver<566 brand=vcs,driver>=565,driver<566 brand=vws,driver>=565,driver<566 brand=cloudgaming,driver>=565,driver<566 brand=unknown,driver>=570,driver<571 brand=grid,driver>=570,driver<571 brand=tesla,driver>=570,driver<571 brand=nvidia,driver>=570,driver<571 brand=quadro,driver>=570,driver<571 brand=quadrortx,driver>=570,driver<571 brand=nvidiartx,driver>=570,driver<571 brand=vapps,driver>=570,driver<571 brand=vpc,driver>=570,driver<571 brand=vcs,driver>=570,driver<571 brand=vws,driver>=570,driver<571 brand=cloudgaming,driver>=570,driver<571
CUDA_VERSION=12.9.1
LD_LIBRARY_PATH=/usr/local/nvidia/lib64:/usr/local/cuda/lib64:/usr/local/cuda/lib64
NVIDIA_DRIVER_CAPABILITIES=compute,utility
VLLM_ENABLE_CUDA_COMPATIBILITY=0
TORCH_CUDA_ARCH_LIST=7.0 7.5 8.0 8.9 9.0 10.0 12.0
VLLM_USAGE_SOURCE=production-docker-image
PYTORCH_NVML_BASED_CUDA_CHECK=1
TORCHINDUCTOR_COMPILE_THREADS=1
TORCHINDUCTOR_CACHE_DIR=/tmp/torchinductor_root

---

# Start vLLM with auto-fit (max_model_len=-1)
  python -m vllm.entrypoints.openai.api_server \
    --model google/gemma-4-31B-it \
    --max-model-len -1 \
    --tensor-parallel-size 1 \
    --gpu-memory-utilization 0.9 \
    --served-model-name vllm-model \
    --port 8080

---

INFO [model.py:1678] Using max model len 262144
  INFO [kv_cache_utils.py:1462] Auto-fit max_model_len: reduced from 262144 to 15168
  INFO [kv_cache_utils.py:1324] Maximum concurrency for 15,168 tokens per request: 1.00x

---

curl -s http://localhost:8080/v1/models | python3 -m json.tool
  # "max_model_len": 262144  ← should be 15168

---

import requests, json

  def test_prompt(n_tokens):
      prompt = "The quick brown fox jumps over the lazy dog. " * (n_tokens // 10)
      r = requests.post("http://localhost:8080/v1/chat/completions", json={
          "model": "vllm-model",
          "messages": [{"role": "user", "content": f"Summarize: {prompt}"}],
          "max_tokens": 16,
      })
      print(f"{n_tokens:>6} tokens: {r.status_code} - {r.json().get('error', {}).get('message', 'OK')[:80]}")

  for n in [4000, 8000, 12000, 15000, 15500, 16000]:
      test_prompt(n)

---

4000 tokens: 200 - OK
    8000 tokens: 200 - OK
   12000 tokens: 200 - OK
   15000 tokens: 200 - OK
   15500 tokens: 500 - EngineCore encountered an issue. See stack trace (above) for the root caus
   16000 tokens: (connection refused — engine is dead, pod restarting)

---

(EngineCore pid=401) AssertionError: Sampled token IDs exceed the max model length. Total number of tokens: 15520 >
  max_model_len: 15168
  (EngineCore pid=401)   File "vllm/v1/engine/core.py", line 1181, in _process_engine_step
  (EngineCore pid=401)     outputs, model_executed = self.step_fn()
  (EngineCore pid=401)   File "vllm/v1/engine/core.py", line 499, in step_with_batch_queue
  (EngineCore pid=401)     model_output = future.result()
  (EngineCore pid=401)   File "vllm/v1/executor/uniproc_executor.py", line 84, in collective_rpc
  (EngineCore pid=401)     result = run_method(self.driver_worker, method, args, kwargs)
  (EngineCore pid=401)   File "vllm/v1/worker/gpu_worker.py", line 740, in sample_tokens
  (EngineCore pid=401)     return self.model_runner.sample_tokens(grammar_output)
  (EngineCore pid=401)   File "vllm/v1/worker/gpu_model_runner.py", line 4282, in sample_tokens
  (EngineCore pid=401)     ) = self._bookkeeping_sync(
  (EngineCore pid=401)   File "vllm/v1/worker/gpu_model_runner.py", line 3446, in _bookkeeping_sync
  (EngineCore pid=401)     assert end_idx <= self.max_model_len, (
  (EngineCore pid=401) AssertionError: Sampled token IDs exceed the max model length. Total number of tokens: 15520 > max_model_len: 15168

  (APIServer pid=1) ERROR [async_llm.py:707] AsyncLLM output_handler failed.
  (APIServer pid=1) vllm.v1.engine.exceptions.EngineDeadError: EngineCore encountered an issue.
  (APIServer pid=1) INFO: Shutting down

RAW_BUFFERClick to expand / collapse

Your current environment

<details> <summary>The output of <code>python collect_env.py</code></summary>

Docker image: vllm/vllm-openai:gemma4
Collecting environment information...
==============================
        System Info
==============================
OS                           : Ubuntu 22.04.5 LTS (x86_64)
GCC version                  : (Ubuntu 11.4.0-1ubuntu1~22.04.3) 11.4.0
Clang version                : Could not collect
CMake version                : Could not collect
Libc version                 : glibc-2.35

==============================
       PyTorch Info
==============================
PyTorch version              : 2.10.0+cu129
Is debug build               : False
CUDA used to build PyTorch   : 12.9
ROCM used to build PyTorch   : N/A
XPU used to build PyTorch    : N/A

==============================
      Python Environment
==============================
Python version               : 3.12.13 (main, Mar  4 2026, 09:23:07) [GCC 11.4.0] (64-bit runtime)
Python platform              : Linux-5.15.0-1074-oracle-x86_64-with-glibc2.35

==============================
       CUDA / GPU Info
==============================
Is CUDA available            : True
CUDA runtime version         : 12.9.86
CUDA_MODULE_LOADING set to   :
GPU models and configuration : GPU 0: NVIDIA H100 80GB HBM3
Nvidia driver version        : 580.126.09
cuDNN version                : Could not collect
HIP runtime version          : N/A
MIOpen runtime version       : N/A
Is XNNPACK available         : True

==============================
          CPU Info
==============================
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):                               224
On-line CPU(s) list:                  0-223
Vendor ID:                            GenuineIntel
Model name:                           Intel(R) Xeon(R) Platinum 8480+
CPU family:                           6
Model:                                143
Thread(s) per core:                   2
Core(s) per socket:                   56
Socket(s):                            2
Stepping:                             8
CPU max MHz:                          3800.0000
CPU min MHz:                          800.0000
BogoMIPS:                             4000.00
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 intel_ppin 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 hwp hwp_act_window hwp_epp hwp_pkg_req 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:                            5.3 MiB (112 instances)
L1i cache:                            3.5 MiB (112 instances)
L2 cache:                             224 MiB (112 instances)
L3 cache:                             210 MiB (2 instances)
NUMA node(s):                         2
NUMA node0 CPU(s):                    0-55,112-167
NUMA node1 CPU(s):                    56-111,168-223
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 Reg file data sampling: Not affected
Vulnerability Retbleed:               Not affected
Vulnerability Spec rstack overflow:   Not affected
Vulnerability Spec store bypass:      Vulnerable
Vulnerability Spectre v1:             Vulnerable: __user pointer sanitization and usercopy barriers only; no swapgs barriers
Vulnerability Spectre v2:             Vulnerable; IBPB: disabled; STIBP: disabled; PBRSB-eIBRS: Vulnerable; BHI: Vulnerable
Vulnerability Srbds:                  Not affected
Vulnerability Tsx async abort:        Not affected

==============================
Versions of relevant libraries
==============================
[pip3] flashinfer-python==0.6.6
[pip3] numpy==2.2.6
[pip3] nvidia-cublas-cu12==12.9.1.4
[pip3] nvidia-cuda-cupti-cu12==12.9.79
[pip3] nvidia-cuda-nvrtc-cu12==12.9.86
[pip3] nvidia-cuda-runtime-cu12==12.9.79
[pip3] nvidia-cudnn-cu12==9.10.2.21
[pip3] nvidia-cudnn-frontend==1.18.0
[pip3] nvidia-cufft-cu12==11.4.1.4
[pip3] nvidia-cufile-cu12==1.14.1.1
[pip3] nvidia-curand-cu12==10.3.10.19
[pip3] nvidia-cusolver-cu12==11.7.5.82
[pip3] nvidia-cusparse-cu12==12.5.10.65
[pip3] nvidia-cusparselt-cu12==0.7.1
[pip3] nvidia-cutlass-dsl==4.4.2
[pip3] nvidia-cutlass-dsl-libs-base==4.4.2
[pip3] nvidia-ml-py==13.595.45
[pip3] nvidia-nccl-cu12==2.27.5
[pip3] nvidia-nvjitlink-cu12==12.9.86
[pip3] nvidia-nvshmem-cu12==3.4.5
[pip3] nvidia-nvtx-cu12==12.9.79
[pip3] pyzmq==27.1.0
[pip3] torch==2.10.0+cu129
[pip3] torch_c_dlpack_ext==0.1.5
[pip3] torchaudio==2.10.0+cu129
[pip3] torchvision==0.25.0+cu129
[pip3] transformers==5.5.0
[pip3] triton==3.6.0
[conda] Could not collect

==============================
         vLLM Info
==============================
ROCM Version                 : Could not collect
vLLM Version                 : 0.19.1.dev6+g6d4a8e6d2 (git sha: 6d4a8e6d2)
vLLM Build Flags:
  CUDA Archs: 7.0 7.5 8.0 8.9 9.0 10.0 12.0; ROCm: Disabled; XPU: Disabled
GPU Topology:
  	GPU0	NIC0	NIC1	NIC2	NIC3	NIC4	NIC5	NIC6	NIC7	NIC8	NIC9	NIC10	NIC11	NIC12	NIC13	NIC14	NIC15	NIC16	NIC17	CPU Affinity	NUMA Affinity	GPU NUMA ID
GPU0	 X 	NODE	NODE	NODE	NODE	NODE	NODE	NODE	PXB	PXB	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	0-55,112-167	0		N/A
NIC0	NODE	 X 	PIX	NODE	NODE	NODE	NODE	NODE	NODE	NODE	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS
NIC1	NODE	PIX	 X 	NODE	NODE	NODE	NODE	NODE	NODE	NODE	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS
NIC2	NODE	NODE	NODE	 X 	NODE	NODE	NODE	NODE	NODE	NODE	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS
NIC3	NODE	NODE	NODE	NODE	 X 	PIX	NODE	NODE	NODE	NODE	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS
NIC4	NODE	NODE	NODE	NODE	PIX	 X 	NODE	NODE	NODE	NODE	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS
NIC5	NODE	NODE	NODE	NODE	NODE	NODE	 X 	PIX	NODE	NODE	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS
NIC6	NODE	NODE	NODE	NODE	NODE	NODE	PIX	 X 	NODE	NODE	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS
NIC7	PXB	NODE	NODE	NODE	NODE	NODE	NODE	NODE	 X 	PIX	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS
NIC8	PXB	NODE	NODE	NODE	NODE	NODE	NODE	NODE	PIX	 X 	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS
NIC9	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	 X 	PIX	NODE	NODE	NODE	NODE	NODE	NODE	NODE
NIC10	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	PIX	 X 	NODE	NODE	NODE	NODE	NODE	NODE	NODE
NIC11	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	NODE	NODE	 X 	NODE	NODE	NODE	NODE	NODE	NODE
NIC12	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	NODE	NODE	NODE	 X 	PIX	NODE	NODE	NODE	NODE
NIC13	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	NODE	NODE	NODE	PIX	 X 	NODE	NODE	NODE	NODE
NIC14	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	NODE	NODE	NODE	NODE	NODE	 X 	PIX	NODE	NODE
NIC15	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	NODE	NODE	NODE	NODE	NODE	PIX	 X 	NODE	NODE
NIC16	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	NODE	NODE	NODE	NODE	NODE	NODE	NODE	 X 	PIX
NIC17	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	SYS	NODE	NODE	NODE	NODE	NODE	NODE	NODE	PIX	 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
  NIC8: mlx5_8
  NIC9: mlx5_9
  NIC10: mlx5_10
  NIC11: mlx5_11
  NIC12: mlx5_12
  NIC13: mlx5_13
  NIC14: mlx5_14
  NIC15: mlx5_15
  NIC16: mlx5_16
  NIC17: mlx5_17

==============================
     Environment Variables
==============================
NVIDIA_VISIBLE_DEVICES=GPU-960f71eb-1b3d-f6d2-bfcb-1a580f51b809
VLLM_RPC_TIMEOUT=30000
VLLM_ENGINE_ITERATION_TIMEOUT_S=120
NVIDIA_REQUIRE_CUDA=cuda>=12.9 brand=unknown,driver>=535,driver<536 brand=grid,driver>=535,driver<536 brand=tesla,driver>=535,driver<536 brand=nvidia,driver>=535,driver<536 brand=quadro,driver>=535,driver<536 brand=quadrortx,driver>=535,driver<536 brand=nvidiartx,driver>=535,driver<536 brand=vapps,driver>=535,driver<536 brand=vpc,driver>=535,driver<536 brand=vcs,driver>=535,driver<536 brand=vws,driver>=535,driver<536 brand=cloudgaming,driver>=535,driver<536 brand=unknown,driver>=550,driver<551 brand=grid,driver>=550,driver<551 brand=tesla,driver>=550,driver<551 brand=nvidia,driver>=550,driver<551 brand=quadro,driver>=550,driver<551 brand=quadrortx,driver>=550,driver<551 brand=nvidiartx,driver>=550,driver<551 brand=vapps,driver>=550,driver<551 brand=vpc,driver>=550,driver<551 brand=vcs,driver>=550,driver<551 brand=vws,driver>=550,driver<551 brand=cloudgaming,driver>=550,driver<551 brand=unknown,driver>=560,driver<561 brand=grid,driver>=560,driver<561 brand=tesla,driver>=560,driver<561 brand=nvidia,driver>=560,driver<561 brand=quadro,driver>=560,driver<561 brand=quadrortx,driver>=560,driver<561 brand=nvidiartx,driver>=560,driver<561 brand=vapps,driver>=560,driver<561 brand=vpc,driver>=560,driver<561 brand=vcs,driver>=560,driver<561 brand=vws,driver>=560,driver<561 brand=cloudgaming,driver>=560,driver<561 brand=unknown,driver>=565,driver<566 brand=grid,driver>=565,driver<566 brand=tesla,driver>=565,driver<566 brand=nvidia,driver>=565,driver<566 brand=quadro,driver>=565,driver<566 brand=quadrortx,driver>=565,driver<566 brand=nvidiartx,driver>=565,driver<566 brand=vapps,driver>=565,driver<566 brand=vpc,driver>=565,driver<566 brand=vcs,driver>=565,driver<566 brand=vws,driver>=565,driver<566 brand=cloudgaming,driver>=565,driver<566 brand=unknown,driver>=570,driver<571 brand=grid,driver>=570,driver<571 brand=tesla,driver>=570,driver<571 brand=nvidia,driver>=570,driver<571 brand=quadro,driver>=570,driver<571 brand=quadrortx,driver>=570,driver<571 brand=nvidiartx,driver>=570,driver<571 brand=vapps,driver>=570,driver<571 brand=vpc,driver>=570,driver<571 brand=vcs,driver>=570,driver<571 brand=vws,driver>=570,driver<571 brand=cloudgaming,driver>=570,driver<571
CUDA_VERSION=12.9.1
LD_LIBRARY_PATH=/usr/local/nvidia/lib64:/usr/local/cuda/lib64:/usr/local/cuda/lib64
NVIDIA_DRIVER_CAPABILITIES=compute,utility
VLLM_ENABLE_CUDA_COMPATIBILITY=0
TORCH_CUDA_ARCH_LIST=7.0 7.5 8.0 8.9 9.0 10.0 12.0
VLLM_USAGE_SOURCE=production-docker-image
PYTORCH_NVML_BASED_CUDA_CHECK=1
TORCHINDUCTOR_COMPILE_THREADS=1
TORCHINDUCTOR_CACHE_DIR=/tmp/torchinductor_root

</details>

🐛 Describe the bug

When --max-model-len=-1 is used (the default auto-fit behavior), vLLM correctly reduces max_model_len from the model config value (262,144) to what fits in GPU memory (15,168 on 1× H100-80GB). However, the admission layer still uses the original 262,144 value, so prompts between 15,168 and 262,144 tokens are accepted into the engine pipeline and then crash it with a fatal AssertionError in _bookkeeping_sync.

Expected behavior

Prompts exceeding the auto-fitted max_model_len should be rejected at admission with HTTP 400 ("prompt too long"). Thw engine should remain alive.

Actual behavior

/v1/models advertises "max_model_len": 262144 (the HF config value, not the auto-fitted 15,168)
A prompt with ~15,500 tokens passes admission
The engine crashes at gpu_model_runner.py:3446 with AssertionError
EngineDeadError propagates → HTTP 500 → process exits → pod restarts (~2-3 min)
All other in-flight requests on that engine get 503

Two sub-issues

/v1/models reports the wrong max_model_len — it should report the auto-fitted value (15,168), not the model config value (262,144). Clients reading this field to self-limit will still send overlong prompts.
No admission rejection — the auto-fitted limit is not propagated to the request validation layer, so overlong prompts slip through to the engine where they hit a fatal assert instead of a graceful 400.

Reproduction

# Start vLLM with auto-fit (max_model_len=-1)
python -m vllm.entrypoints.openai.api_server \
  --model google/gemma-4-31B-it \
  --max-model-len -1 \
  --tensor-parallel-size 1 \
  --gpu-memory-utilization 0.9 \
  --served-model-name vllm-model \
  --port 8080

Observe in startup logs:

INFO [model.py:1678] Using max model len 262144
INFO [kv_cache_utils.py:1462] Auto-fit max_model_len: reduced from 262144 to 15168
INFO [kv_cache_utils.py:1324] Maximum concurrency for 15,168 tokens per request: 1.00x

Then verify /v1/models reports 262144 (wrong):

curl -s http://localhost:8080/v1/models | python3 -m json.tool
# "max_model_len": 262144  ← should be 15168

Send prompts of increasing size:

import requests, json

def test_prompt(n_tokens):
    prompt = "The quick brown fox jumps over the lazy dog. " * (n_tokens // 10)
    r = requests.post("http://localhost:8080/v1/chat/completions", json={
        "model": "vllm-model",
        "messages": [{"role": "user", "content": f"Summarize: {prompt}"}],
        "max_tokens": 16,
    })
    print(f"{n_tokens:>6} tokens: {r.status_code} - {r.json().get('error', {}).get('message', 'OK')[:80]}")

for n in [4000, 8000, 12000, 15000, 15500, 16000]:
    test_prompt(n)

Results:

  4000 tokens: 200 - OK
  8000 tokens: 200 - OK
 12000 tokens: 200 - OK
 15000 tokens: 200 - OK
 15500 tokens: 500 - EngineCore encountered an issue. See stack trace (above) for the root caus
 16000 tokens: (connection refused — engine is dead, pod restarting)

Stack trace

(EngineCore pid=401) AssertionError: Sampled token IDs exceed the max model length. Total number of tokens: 15520 >
max_model_len: 15168
(EngineCore pid=401)   File "vllm/v1/engine/core.py", line 1181, in _process_engine_step
(EngineCore pid=401)     outputs, model_executed = self.step_fn()
(EngineCore pid=401)   File "vllm/v1/engine/core.py", line 499, in step_with_batch_queue
(EngineCore pid=401)     model_output = future.result()
(EngineCore pid=401)   File "vllm/v1/executor/uniproc_executor.py", line 84, in collective_rpc
(EngineCore pid=401)     result = run_method(self.driver_worker, method, args, kwargs)
(EngineCore pid=401)   File "vllm/v1/worker/gpu_worker.py", line 740, in sample_tokens
(EngineCore pid=401)     return self.model_runner.sample_tokens(grammar_output)
(EngineCore pid=401)   File "vllm/v1/worker/gpu_model_runner.py", line 4282, in sample_tokens
(EngineCore pid=401)     ) = self._bookkeeping_sync(
(EngineCore pid=401)   File "vllm/v1/worker/gpu_model_runner.py", line 3446, in _bookkeeping_sync
(EngineCore pid=401)     assert end_idx <= self.max_model_len, (
(EngineCore pid=401) AssertionError: Sampled token IDs exceed the max model length. Total number of tokens: 15520 > max_model_len: 15168

(APIServer pid=1) ERROR [async_llm.py:707] AsyncLLM output_handler failed.
(APIServer pid=1) vllm.v1.engine.exceptions.EngineDeadError: EngineCore encountered an issue.
(APIServer pid=1) INFO: Shutting down

Before submitting a new issue...

Make sure you already searched for relevant issues, and asked the chatbot living at the bottom right corner of the documentation page, which can answer lots of frequently asked questions.

extent analysis

TL;DR

The issue can be fixed by updating the max_model_len value reported by the /v1/models endpoint to reflect the auto-fitted value and by adding admission rejection for prompts exceeding this limit.

Guidance

Update the /v1/models endpoint: Modify the code to report the auto-fitted max_model_len value (15,168) instead of the model config value (262,144).
Add admission rejection: Implement a check to reject prompts that exceed the auto-fitted max_model_len value, returning an HTTP 400 error ("prompt too long").
Verify the fix: Test the updated endpoint and admission layer to ensure that prompts within the auto-fitted limit are accepted and those exceeding it are rejected with an HTTP 400 error.
Review the stack trace: Examine the provided stack trace to understand the flow of the error and ensure that the proposed fix addresses the root cause of the issue.

Example

No code snippet is provided as the issue description does not include specific code that needs to be modified. However, the fix would involve updating the relevant parts of the vllm codebase to reflect the auto-fitted max_model_len value and adding a prompt length check.

Notes

The provided information suggests that the issue is specific to the vllm system and its interaction with GPU memory and model configuration. The fix should be applied carefully to avoid introducing other issues or affecting the system's performance.

Recommendation

Apply the workaround by updating the /v1/models endpoint and adding admission rejection for prompts exceeding the auto-fitted max_model_len value. This should prevent the engine from crashing due to prompts that are too long and provide a more robust user experience.

Vote matrix · Quick signals

Works

Did the solution work? Tap to confirm.

Easy Fix

Was it a quick fix?

Time Saver

Did it save you time?

Blocking

Was it severely blocking?

Common Issue

Are others likely hitting this too?

Flaky / Intermittent

Is it intermittent?

Verified / Reproducible

Can you reproduce it reliably?

#api #tokenizer error #prompt formatting #chain error #environment variable

Still need to ship something?

×6

Another batch ranked right after the header list — different links, same matching logic.

Data

Security

Network

Code

UI/UX

Text

System

Multimedia

Protocol

API

Engineering

vllm - 💡(How to fix) Fix [Bug]: Engine crashes with AssertionError when prompt exceeds auto-fitted max_model_len (admission check missing) [1 comments, 2 participants]

Recommended Tools

GitHub issue graph ai analysis

Error Message

Code Example

Your current environment

🐛 Describe the bug

Expected behavior

Actual behavior

Two sub-issues

Reproduction

Stack trace

Before submitting a new issue...

extent analysis

TL;DR

Guidance

Example

Notes

Recommendation

Still need to ship something?

TRENDING

vllm - 💡(How to fix) Fix [Bug]: Engine crashes with AssertionError when prompt exceeds auto-fitted max_model_len (admission check missing) [1 comments, 2 participants]

Recommended Tools

GitHub issue graph ai analysis

Error Message

Code Example

Your current environment

🐛 Describe the bug

Expected behavior

Actual behavior

Two sub-issues

Reproduction

Stack trace

Before submitting a new issue...

extent analysis

TL;DR

Guidance

Example

Notes

Recommendation

Still need to ship something?

RELATED_DISCOVERY

TRENDING