vllm: [Bug]: 2 nodes serving hanging
Your current environment
The output of `python collect_env.py`
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 20.04.6 LTS (x86_64)
GCC version: (Ubuntu 9.4.0-1ubuntu1~20.04.2) 9.4.0
Clang version: Could not collect
CMake version: Could not collect
Libc version: glibc-2.31
Python version: 3.10.14 (main, Apr 6 2024, 18:45:05) [GCC 9.4.0] (64-bit runtime)
Python platform: Linux-5.4.119-19.0009.28-x86_64-with-glibc2.31
Is CUDA available: True
CUDA runtime version: Could not collect
CUDA_MODULE_LOADING set to: LAZY
GPU models and configuration:
GPU 0: NVIDIA H800
GPU 1: NVIDIA H800
GPU 2: NVIDIA H800
GPU 3: NVIDIA H800
GPU 4: NVIDIA H800
GPU 5: NVIDIA H800
GPU 6: NVIDIA H800
GPU 7: NVIDIA H800
Nvidia driver version: 535.54.03
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
Byte Order: Little Endian
Address sizes: 46 bits physical, 48 bits virtual
CPU(s): 200
On-line CPU(s) list: 0-199
Thread(s) per core: 2
Core(s) per socket: 50
Socket(s): 2
NUMA node(s): 2
Vendor ID: GenuineIntel
CPU family: 6
Model: 143
Model name: Intel(R) Xeon(R) Platinum 8480+
Stepping: 6
CPU MHz: 2000.000
BogoMIPS: 4000.00
Hypervisor vendor: KVM
Virtualization type: full
L1d cache: 4.7 MiB
L1i cache: 3.1 MiB
L2 cache: 200 MiB
L3 cache: 210 MiB
NUMA node0 CPU(s): 0-99
NUMA node1 CPU(s): 100-199
Vulnerability Itlb multihit: Not affected
Vulnerability L1tf: Not affected
Vulnerability Mds: Not affected
Vulnerability Meltdown: 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
Vulnerability Srbds: Not affected
Vulnerability Tsx async abort: Not affected
Flags: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ss ht syscall nx pdpe1gb rdtscp lm constant_tsc rep_good nopl xtopology nonstop_tsc cpuid tsc_known_freq pni pclmulqdq monitor ssse3 fma cx16 pcid sse4_1 sse4_2 x2apic movbe popcnt tsc_deadline_timer aes xsave avx f16c rdrand hypervisor lahf_lm abm 3dnowprefetch cpuid_fault invpcid_single ssbd ibrs ibpb ibrs_enhanced fsgsbase bmi1 hle avx2 smep bmi2 erms invpcid rtm avx512f avx512dq rdseed adx smap avx512ifma clflushopt clwb avx512cd sha_ni avx512bw avx512vl xsaveopt xsavec xgetbv1 xsaves avx512_bf16 wbnoinvd arat avx512vbmi umip avx512_vbmi2 gfni vaes vpclmulqdq avx512_vnni avx512_bitalg avx512_vpopcntdq movdiri movdir64b fsrm arch_capabilities
Versions of relevant libraries:
[pip3] flashinfer==0.1.6+cu121torch2.4
[pip3] numpy==1.26.4
[pip3] nvidia-cublas-cu12==12.1.3.1
[pip3] nvidia-cuda-cupti-cu12==12.1.105
[pip3] nvidia-cuda-nvrtc-cu12==12.1.105
[pip3] nvidia-cuda-runtime-cu12==12.1.105
[pip3] nvidia-cudnn-cu12==9.1.0.70
[pip3] nvidia-cufft-cu12==11.0.2.54
[pip3] nvidia-curand-cu12==10.3.2.106
[pip3] nvidia-cusolver-cu12==11.4.5.107
[pip3] nvidia-cusparse-cu12==12.1.0.106
[pip3] nvidia-ml-py==12.560.30
[pip3] nvidia-nccl-cu12==2.20.5
[pip3] nvidia-nvjitlink-cu12==12.6.68
[pip3] nvidia-nvtx-cu12==12.1.105
[pip3] pyzmq==26.2.0
[pip3] torch==2.4.0
[pip3] torchvision==0.19.0
[pip3] transformers==4.44.2
[pip3] triton==3.0.0
[conda] Could not collect
ROCM Version: Could not collect
Neuron SDK Version: N/A
vLLM Version: 0.6.0@COMMIT_HASH_PLACEHOLDER
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 NV8 NV8 NV8 NV8 NV8 NV8 NV8 PIX NODE NODE NODE SYS SYS SYS SYS 0-99 0 N/A
GPU1 NV8 X NV8 NV8 NV8 NV8 NV8 NV8 NODE PIX NODE NODE SYS SYS SYS SYS 0-99 0 N/A
GPU2 NV8 NV8 X NV8 NV8 NV8 NV8 NV8 NODE NODE PIX NODE SYS SYS SYS SYS 0-99 0 N/A
GPU3 NV8 NV8 NV8 X NV8 NV8 NV8 NV8 NODE NODE NODE PIX SYS SYS SYS SYS 0-99 0 N/A
GPU4 NV8 NV8 NV8 NV8 X NV8 NV8 NV8 SYS SYS SYS SYS PIX NODE NODE NODE 100-199 1 N/A
GPU5 NV8 NV8 NV8 NV8 NV8 X NV8 NV8 SYS SYS SYS SYS NODE PIX NODE NODE 100-199 1 N/A
GPU6 NV8 NV8 NV8 NV8 NV8 NV8 X NV8 SYS SYS SYS SYS NODE NODE PIX NODE 100-199 1 N/A
GPU7 NV8 NV8 NV8 NV8 NV8 NV8 NV8 X SYS SYS SYS SYS NODE NODE NODE PIX 100-199 1 N/A
NIC0 PIX NODE NODE NODE SYS SYS SYS SYS X NODE NODE NODE SYS SYS SYS SYS
NIC1 NODE PIX NODE NODE SYS SYS SYS SYS NODE X NODE NODE SYS SYS SYS SYS
NIC2 NODE NODE PIX NODE SYS SYS SYS SYS NODE NODE X NODE SYS SYS SYS SYS
NIC3 NODE NODE NODE PIX SYS SYS SYS SYS NODE NODE NODE X SYS SYS SYS SYS
NIC4 SYS SYS SYS SYS PIX NODE NODE NODE SYS SYS SYS SYS X NODE NODE NODE
NIC5 SYS SYS SYS SYS NODE PIX NODE NODE SYS SYS SYS SYS NODE X NODE NODE
NIC6 SYS SYS SYS SYS NODE NODE PIX NODE SYS SYS SYS SYS NODE NODE X NODE
NIC7 SYS SYS SYS SYS NODE NODE NODE PIX SYS SYS SYS SYS NODE NODE NODE 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_bond_0
NIC1: mlx5_bond_1
NIC2: mlx5_bond_2
NIC3: mlx5_bond_3
NIC4: mlx5_bond_4
NIC5: mlx5_bond_5
NIC6: mlx5_bond_6
NIC7: mlx5_bond_7
Model Input Dumps
No response
🐛 Describe the bug
running command:
export NCCL_DEBUG=INFO export CUDA_LAUNCH_BLOCKING=1 export VLLM_TRACE_FUNCTION=1
python3 api_server.py
–model my_model
-tp 8
-pp 2
–enforce-eager
–max-num-seqs=32
–dtype=bfloat16
–worker-use-ray
–gpu-memory-utilization 0.8
logs are hanging here:
INFO 09-12 23:40:47 utils.py:977] Found nccl from library libnccl.so.2 INFO 09-12 23:40:47 pynccl.py:63] vLLM is using nccl==2.20.5 VM-160-69-tencentos:561:561 [0] NCCL INFO Using non-device net plugin version 0 VM-160-69-tencentos:561:561 [0] NCCL INFO Using network IB VM-160-69-tencentos:561:561 [0] NCCL INFO comm 0x1476dca0 rank 0 nranks 2 cudaDev 0 nvmlDev 0 busId 23000 commId 0xa015ae7e0a50005d - Init START (RayWorkerWrapper pid=248, ip=10.1.160.68) INFO 09-12 23:40:47 utils.py:977] Found nccl from library libnccl.so.2 (RayWorkerWrapper pid=248, ip=10.1.160.68) INFO 09-12 23:40:47 pynccl.py:63] vLLM is using nccl==2.20.5 VM-160-69-tencentos:561:561 [0] NCCL INFO Setting affinity for GPU 0 to 0f,ffffffff,ffffffff,ffffffff VM-160-69-tencentos:561:561 [0] NCCL INFO comm 0x1476dca0 rank 0 nRanks 2 nNodes 2 localRanks 1 localRank 0 MNNVL 0 VM-160-69-tencentos:561:561 [0] NCCL INFO Channel 00/02 : 0 1 VM-160-69-tencentos:561:561 [0] NCCL INFO Channel 01/02 : 0 1 VM-160-69-tencentos:561:561 [0] NCCL INFO Trees [0] 1/-1/-1->0->-1 [1] -1/-1/-1->0->1 VM-160-69-tencentos:561:561 [0] NCCL INFO P2P Chunksize set to 131072 VM-160-69-tencentos:561:561 [0] NCCL INFO Channel 00/0 : 1[0] -> 0[0] [receive] via NET/IB/1/GDRDMA VM-160-69-tencentos:561:561 [0] NCCL INFO Channel 01/0 : 1[0] -> 0[0] [receive] via NET/IB/1/GDRDMA VM-160-69-tencentos:561:561 [0] NCCL INFO Channel 00/0 : 0[0] -> 1[0] [send] via NET/IB/1/GDRDMA VM-160-69-tencentos:561:561 [0] NCCL INFO Channel 01/0 : 0[0] -> 1[0] [send] via NET/IB/1/GDRDMA VM-160-69-tencentos:561:561 [0] NCCL INFO Connected all rings VM-160-69-tencentos:561:561 [0] NCCL INFO Connected all trees VM-160-69-tencentos:561:561 [0] NCCL INFO threadThresholds 8/8/64 | 16/8/64 | 512 | 512 VM-160-69-tencentos:561:561 [0] NCCL INFO 2 coll channels, 0 collnet channels, 0 nvls channels, 2 p2p channels, 2 p2p channels per peer VM-160-69-tencentos:561:561 [0] NCCL INFO comm 0x1476dca0 rank 0 nranks 2 cudaDev 0 nvmlDev 0 busId 23000 commId 0xa015ae7e0a50005d - Init COMPLETE
VM-160-69-tencentos:561:13506 [0] transport/net_ib.cc:1698 NCCL WARN NET/IB : Got completion from peer 10.1.160.68<57552> with status=12 opcode=0 len=0 vendor err 129 (Recv) localGid fe80::a288:c2ff:fe16:5c9c remoteGidsfe80::a288:c2ff:fe16:4e4c VM-160-69-tencentos:561:13506 [0] NCCL INFO transport/net.cc:1298 -> 6 VM-160-69-tencentos:561:13506 [0] NCCL INFO proxy.cc:694 -> 6 VM-160-69-tencentos:561:13506 [0] NCCL INFO proxy.cc:874 -> 6 [Progress Thread]
VM-160-69-tencentos:561:13506 [0] transport/net_ib.cc:1698 NCCL WARN NET/IB : Got completion from peer 10.1.160.68<57552> with status=5 opcode=0 len=0 vendor err 249 (Recv) localGid fe80::a288:c2ff:fe16:5c9c remoteGidsfe80::a288:c2ff:fe16:4e4c VM-160-69-tencentos:561:13506 [0] NCCL INFO transport/net.cc:1298 -> 6 VM-160-69-tencentos:561:13506 [0] NCCL INFO proxy.cc:694 -> 6 VM-160-69-tencentos:561:13506 [0] NCCL INFO proxy.cc:874 -> 6 [Progress Thread]
VM-160-69-tencentos:561:13506 [0] transport/net_ib.cc:1698 NCCL WARN NET/IB : Got completion from peer 10.1.160.68<57552> with status=5 opcode=0 len=0 vendor err 244 (Recv) localGid fe80::a288:c2ff:fe16:5c9c remoteGidsfe80::a288:c2ff:fe16:4e4c VM-160-69-tencentos:561:13506 [0] NCCL INFO transport/net.cc:1298 -> 6 VM-160-69-tencentos:561:13506 [0] NCCL INFO proxy.cc:694 -> 6 VM-160-69-tencentos:561:13506 [0] NCCL INFO proxy.cc:874 -> 6 [Progress Thread]
VM-160-69-tencentos:561:13506 [0] transport/net_ib.cc:1698 NCCL WARN NET/IB : Got completion from peer 10.1.160.68<57552> with status=5 opcode=0 len=0 vendor err 249 (Recv) localGid fe80::a288:c2ff:fe16:5c9c remoteGidsfe80::a288:c2ff:fe16:4e4c VM-160-69-tencentos:561:13506 [0] NCCL INFO transport/net.cc:1298 -> 6 VM-160-69-tencentos:561:13506 [0] NCCL INFO proxy.cc:694 -> 6 VM-160-69-tencentos:561:13506 [0] NCCL INFO proxy.cc:874 -> 6 [Progress Thread]
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.
About this issue
- Original URL
- State: closed
- Created 2 years ago
- Comments: 19 (9 by maintainers)
export NCCL_IB_GID_INDEX=3 fix this (default setting is GID 0/1 and IPV6 is not setting correctly), thank you!
node1:
node 1 got all_reduce_sum result=0, strange.
NCCL_DEBUG=TRACE torchrun --nnodes 2 --nproc-per-node=8 --node-rank 0 --master_addr 10.1.160.68 --master_port 23456 test.py
global world size = 16, however:
node0:
node1:
