Skip to content

Commit 7378832

Browse files
Create test-native-gpu-runner.yaml
1 parent bb86261 commit 7378832

File tree

1 file changed

+72
-0
lines changed

1 file changed

+72
-0
lines changed
Lines changed: 72 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,72 @@
1+
name: Test - Native GPU Runner
2+
3+
on:
4+
workflow_dispatch:
5+
push:
6+
paths:
7+
- ".github/workflows/test-native-gpu-runner.yaml"
8+
9+
jobs:
10+
test-2gpu-native:
11+
runs-on: g6-2gpu-native-runner
12+
steps:
13+
- name: Checkout
14+
uses: actions/checkout@v4
15+
16+
- name: Job Info
17+
run: |
18+
echo "=== Native GPU Runner Test ==="
19+
echo "Runner: $(hostname)"
20+
echo "Timestamp: $(date -u)"
21+
22+
- name: Check GPU Environment
23+
run: |
24+
echo "=== GPU Environment ==="
25+
echo "NVIDIA_VISIBLE_DEVICES: ${NVIDIA_VISIBLE_DEVICES:-not set}"
26+
echo "CUDA_VISIBLE_DEVICES: ${CUDA_VISIBLE_DEVICES:-not set}"
27+
28+
- name: Check GPU Info
29+
run: |
30+
echo "=== GPU Information ==="
31+
nvidia-smi
32+
nvidia-smi -L
33+
nvidia-smi --query-gpu=index,name,uuid,memory.total,memory.used --format=csv
34+
35+
- name: Check CUDA
36+
run: |
37+
echo "=== CUDA Info ==="
38+
nvcc --version || echo "nvcc not in PATH"
39+
ls -la /usr/local/cuda/bin/ || true
40+
41+
- name: Check Node Resources
42+
run: |
43+
echo "=== Node Information ==="
44+
echo "Hostname: $(hostname)"
45+
echo "CPU cores: $(nproc)"
46+
echo "Memory: $(free -h | grep Mem | awk '{print $2}')"
47+
48+
# Test parallel jobs to verify GPU isolation
49+
test-parallel-native:
50+
strategy:
51+
matrix:
52+
job_id: [1, 2]
53+
runs-on: g6-2gpu-native-runner
54+
steps:
55+
- name: Job Info
56+
run: |
57+
echo "=== Parallel Native GPU Test ==="
58+
echo "Job ID: ${{ matrix.job_id }}"
59+
echo "Hostname: $(hostname)"
60+
61+
- name: Check GPU Allocation
62+
run: |
63+
echo "=== GPU Allocation Check ==="
64+
echo "NVIDIA_VISIBLE_DEVICES: ${NVIDIA_VISIBLE_DEVICES:-not set}"
65+
nvidia-smi -L
66+
nvidia-smi --query-gpu=index,uuid --format=csv
67+
68+
- name: Simulate Workload
69+
run: |
70+
echo "=== Simulating GPU Workload ==="
71+
sleep 30
72+
echo "✅ Job ${{ matrix.job_id }} completed"

0 commit comments

Comments
 (0)