[Bug]: Multi GPU setup for VLLM in Openshift still does not work #5360
Description
Your current environment
Collecting environment information...
PyTorch version: 2.3.0+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 3.29.3
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-5.14.0-284.66.1.el9_2.x86_64-x86_64-with-glibc2.35
Is CUDA available: True
CUDA runtime version: Could not collect
CUDA_MODULE_LOADING set to: LAZY
GPU models and configuration:
GPU 0: NVIDIA L40S
GPU 1: NVIDIA L40S
GPU 2: NVIDIA L40S
GPU 3: NVIDIA L40S
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: 52 bits physical, 57 bits virtual
Byte Order: Little Endian
CPU(s): 128
On-line CPU(s) list: 0-127
Vendor ID: AuthenticAMD
Model name: AMD EPYC 9334 32-Core Processor
CPU family: 25
Model: 17
Thread(s) per core: 2
Core(s) per socket: 32
Socket(s): 2
Stepping: 1
Frequency boost: enabled
CPU max MHz: 3910.2529
CPU min MHz: 1500.0000
BogoMIPS: 5400.11
Flags: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm constant_tsc rep_good nopl nonstop_tsc cpuid extd_apicid aperfmperf rapl pni pclmulqdq monitor ssse3 fma cx16 pcid sse4_1 sse4_2 x2apic movbe popcnt aes xsave avx f16c rdrand lahf_lm cmp_legacy svm extapic cr8_legacy abm sse4a misalignsse 3dnowprefetch osvw ibs skinit wdt tce topoext perfctr_core perfctr_nb bpext perfctr_llc mwaitx cpb cat_l3 cdp_l3 invpcid_single hw_pstate ssbd mba perfmon_v2 ibrs ibpb stibp vmmcall fsgsbase bmi1 avx2 smep bmi2 erms invpcid cqm rdt_a avx512f avx512dq rdseed adx smap avx512ifma clflushopt clwb avx512cd sha_ni avx512bw avx512vl xsaveopt xsavec xgetbv1 xsaves cqm_llc cqm_occup_llc cqm_mbm_total cqm_mbm_local avx512_bf16 clzero irperf xsaveerptr rdpru wbnoinvd amd_ppin cppc arat npt lbrv svm_lock nrip_save tsc_scale vmcb_clean flushbyasid decodeassists pausefilter pfthreshold avic v_vmsave_vmload vgif x2avic v_spec_ctrl avx512vbmi umip pku ospke avx512_vbmi2 gfni vaes vpclmulqdq avx512_vnni avx512_bitalg avx512_vpopcntdq la57 rdpid overflow_recov succor smca fsrm flush_l1d
Virtualization: AMD-V
L1d cache: 2 MiB (64 instances)
L1i cache: 2 MiB (64 instances)
L2 cache: 64 MiB (64 instances)
L3 cache: 256 MiB (8 instances)
NUMA node(s): 2
NUMA node0 CPU(s): 0-31,64-95
NUMA node1 CPU(s): 32-63,96-127
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: Mitigation; Safe RET
Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1: Mitigation; usercopy/swapgs barriers and __user pointer sanitization
Vulnerability Spectre v2: Mitigation; Retpolines, IBPB conditional, IBRS_FW, STIBP always-on, RSB filling, PBRSB-eIBRS Not affected
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==2.3.0
[pip3] transformers==4.41.2
[pip3] triton==2.3.0
[conda] Could not collect
ROCM Version: Could not collect
Neuron SDK Version: N/A
vLLM Version: 0.4.3
vLLM Build Flags:
CUDA Archs: Not Set; ROCm: Disabled; Neuron: Disabled
GPU Topology:
GPU0 GPU1 GPU2 GPU3 NIC0 NIC1 CPU Affinity NUMA Affinity GPU NUMA ID
GPU0 X SYS SYS SYS SYS SYS 0-31,64-95 0 N/A
GPU1 SYS X SYS SYS SYS SYS 0-31,64-95 0 N/A
GPU2 SYS SYS X SYS SYS SYS 32-63,96-127 1 N/A
GPU3 SYS SYS SYS X SYS SYS 32-63,96-127 1 N/A
NIC0 SYS SYS SYS SYS X PIX
NIC1 SYS SYS SYS SYS 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
🐛 Describe the bug
Reposting #4462 as it is still an on going issue.
Vllm best case it inconsistent on if It can start a multi gpu instance within an openshift/k8s enviornment on startup but 99% of the time it fails the start.
Ideally if the Nvidia operator is installed and working correctly then when pasted into the POD, vllm should be able to identify the givens GPUs and start; however, it tends to freeze up on start up at the point
INFO 06-09 00:02:22 llm_engine.py:161] Initializing an LLM engine (v0.4.3) with config: model='mistralai/Mixtral-8x7B-Instruct-v0.1', speculative_config=None, tokenizer='mistralai/Mixtral-8x7B-Instruct-v0.1', skip_tokenizer_init=False, tokenizer_mode=auto, revision=None, rope_scaling=None, tokenizer_revision=None, trust_remote_code=False, dtype=torch.bfloat16, max_seq_len=32768, download_dir=None, load_format=LoadFormat.AUTO, tensor_parallel_size=4, disable_custom_all_reduce=False, quantization=None, enforce_eager=False, kv_cache_dtype=auto, quantization_param_path=None, device_config=cuda, decoding_config=DecodingConfig(guided_decoding_backend='outlines'), seed=0, served_model_name=mistralai/Mixtral-8x7B-Instruct-v0.1)
(VllmWorkerProcess pid=76) INFO 06-09 00:02:25 multiproc_worker_utils.py:214] Worker ready; awaiting tasks
(VllmWorkerProcess pid=77) INFO 06-09 00:02:25 multiproc_worker_utils.py:214] Worker ready; awaiting tasks
(VllmWorkerProcess pid=75) INFO 06-09 00:02:25 multiproc_worker_utils.py:214] Worker ready; awaiting tasks
Here is the yaml file
apiVersion: apps/v1
kind: Deployment
metadata:
name: mixtral-8x7b-instruct-tgi-deploy
labels:
app: mixtral-8x7b-instruct-tgi-deploy
spec:
replicas: 1
revisionHistoryLimit: 1
strategy:
type: Recreate
selector:
matchLabels:
app: mixtral-8x7b-instruct-tgi-pod
template:
metadata:
labels:
app: mixtral-8x7b-instruct-tgi-pod
spec:
volumes:
- name: model
persistentVolumeClaim:
claimName: hub-pv-filesystem
- name: dshm
emptyDir:
medium: Memory
sizeLimit: "15Gi"
containers:
- name: mixtral-8x7b-instruct-tgi-pod
image: vllm/vllm-openai:v0.4.3
args: ["--model mistralai/Mixtral-8x7B-Instruct-v0.1 --gpu-memory-utilization 0.95 --tensor-parallel-size 4 --distributed-executor-backend mp"]
ports:
- containerPort: 8000
readinessProbe:
tcpSocket:
port: 8000
initialDelaySeconds: 30
periodSeconds: 30
resources:
limits:
nvidia.com/gpu: "4"
requests:
cpu: 4
memory: 8Gi
nvidia.com/gpu: "4"
volumeMounts:
- mountPath: /root/.cache/huggingface/hub
name: model
- name: dshm
mountPath: /dev/shm
env:
- name: HUGGING_FACE_HUB_TOKEN
value: xxxxxx
compared to the previous post, I switched the distributed-executor-backend to mp and it still has the same problem
I am running on a L40s.