vllm-ascend/tests/e2e/nightly/multi_node/scripts/lws-a2.yaml.jinja2 at 74e2b71821c00a382f6b62bf8c0d6da32a5305e7 · vllm-project/vllm-ascend · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
apiVersion: leaderworkerset.x-k8s.io/v1
kind: LeaderWorkerSet
metadata:
  name: {{ lws_name | default("vllm") }}
  namespace: vllm-project
spec:
  replicas: {{ replicas | default(1) }}
  leaderWorkerTemplate:
    size: {{ size | default(2) }}
    restartPolicy: None
    leaderTemplate:
      metadata:
        labels:
          role: leader
      spec:
        containers:
          - name: vllm-leader
            imagePullPolicy: Always
            image: {{ image | default("swr.cn-southwest-2.myhuaweicloud.com/base_image/ascend-ci/vllm-ascend:nightly-" + vllm_version + "-a2") }}
            env:
              - name: CONFIG_YAML_PATH
                value: {{ config_file_path | default("DeepSeek-V3.yaml") }}
              - name: WORKSPACE
                value: "/vllm-workspace"
              - name: FAIL_TAG
                value: {{ fail_tag | default("FAIL_TAG") }}
              - name: IS_PR_TEST
                value: "{{ is_pr_test | default("false") }}"
              - name: VLLM_VERSION
                value: {{ vllm_version | default("latest") }}
              - name: VLLM_ASCEND_REF
                value: {{ vllm_ascend_ref | default("main") }}
              - name: VLLM_ASCEND_REMOTE_URL
                value: {{ vllm_ascend_remote_url | default("https://github.com/vllm-project/vllm-ascend.git") }}
            command:
              - sh
              - -c
              - |
                bash /root/.cache/tests/run.sh
            resources:
              limits:
                huawei.com/ascend-1980: {{ npu_per_node | default("16") }}
                memory: 512Gi
                ephemeral-storage: 100Gi
              requests:
                huawei.com/ascend-1980: {{ npu_per_node | default("16") }}
                ephemeral-storage: 100Gi
                cpu: 125
            ports:
              - containerPort: 8080
            # readinessProbe:
            #   tcpSocket:
            #     port: 8080
            #   initialDelaySeconds: 15
            #   periodSeconds: 10
            volumeMounts:
              - mountPath: /root/.cache
                name: shared-volume
              - mountPath: /usr/local/Ascend/driver/tools
                name: driver-tools
              - mountPath: /dev/shm
                name: dshm
        volumes:
        - name: dshm
          emptyDir:
            medium: Memory
            sizeLimit: 15Gi
        - name: shared-volume
          persistentVolumeClaim:
            claimName: vllm-project-hk001
        - name: driver-tools
          hostPath:
            path: /usr/local/Ascend/driver/tools
    workerTemplate:
      spec:
        containers:
          - name: vllm-worker
            imagePullPolicy: Always
            image: {{ image | default("swr.cn-southwest-2.myhuaweicloud.com/base_image/ascend-ci/vllm-ascend:nightly-" + vllm_version + "-a2") }}
            env:
              - name: CONFIG_YAML_PATH
                value: {{ config_file_path | default("DeepSeek-V3.yaml") }}
              - name: WORKSPACE
                value: "/vllm-workspace"
              - name: FAIL_TAG
                value: {{ fail_tag | default("FAIL_TAG") }}
              - name: IS_PR_TEST
                value: "{{ is_pr_test | default("false") }}"
              - name: VLLM_VERSION
                value: {{ vllm_version | default("latest") }}
              - name: VLLM_ASCEND_REF
                value: {{ vllm_ascend_ref | default("main") }}
              - name: VLLM_ASCEND_REMOTE_URL
                value: {{ vllm_ascend_remote_url | default("https://github.com/vllm-project/vllm-ascend.git") }}
            command:
              - sh
              - -c
              - |
                bash /root/.cache/tests/run.sh
            resources:
              limits:
                huawei.com/ascend-1980: {{ npu_per_node | default("16") }}
                memory: 512Gi
                ephemeral-storage: 100Gi
              requests:
                huawei.com/ascend-1980: {{ npu_per_node | default("16") }}
                ephemeral-storage: 100Gi
                cpu: 125
            volumeMounts:
              - mountPath: /root/.cache
                name: shared-volume
              - mountPath: /usr/local/Ascend/driver/tools
                name: driver-tools
              - mountPath: /dev/shm
                name: dshm
        volumes:
        - name: dshm
          emptyDir:
            medium: Memory
            sizeLimit: 15Gi
        - name: shared-volume
          persistentVolumeClaim:
            claimName: vllm-project-hk001
        - name: driver-tools
          hostPath:
            path: /usr/local/Ascend/driver/tools
---
apiVersion: v1
kind: Service
metadata:
  name: {{ lws_name | default("vllm") }}-leader
  namespace: vllm-project
spec:
  ports:
    - name: http
      port: 8080
      protocol: TCP
      targetPort: 8080
  selector:
    leaderworkerset.sigs.k8s.io/name: {{ lws_name | default("vllm") }}
    role: leader
  type: ClusterIP