ivanwu/helm-charts
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

feat: support Hami on k3s and k8s
All checks were successful
Publish Helm Charts / helm-publish (push) Successful in 7s
Details

Browse Source
This commit is contained in:
Ivan087
2025-12-05 17:09:01 +08:00
parent 97c5d559e3
commit 89bc94a6a9
5 changed files with 27 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

14
vllm-serve/templates/lws.yaml
View File

@ -16,7 +16,7 @@ spec:
initContainers:
# 模型下载作为第一个 initContainer
- name: download-model
image: {{ .Values.model.download.image }}
image: alpine:latest
imagePullPolicy: {{ .Values.imagePullPolicy | default "IfNotPresent" }}
env:
- name: HF_ENDPOINT
@ -122,6 +122,9 @@ spec:
nvidia.com/gpu: "{{ .Values.resources.gpuLimit }}"
memory: {{ .Values.resources.memoryLimit }}
ephemeral-storage: 10Gi
{{- if and .Values.resources.gpuMem (gt (int .Values.resources.gpuMem) 0) }}
nvidia.com/gpumem: {{ .Values.resources.gpuMem }}
{{- end }}
{{- if .Values.rdma.enabled }}
{{ .Values.rdma.resourceName }}: {{ .Values.rdma.resourceCount | default 1 }}
{{- end }}
@ -129,6 +132,9 @@ spec:
ephemeral-storage: 10Gi
cpu: {{ .Values.resources.cpuRequest }}
nvidia.com/gpu: "{{ .Values.resources.gpuLimit }}"
{{- if and .Values.resources.gpuMem (gt (int .Values.resources.gpuMem) 0) }}
nvidia.com/gpumem: {{ .Values.resources.gpuMem }}
{{- end }}
{{- if .Values.rdma.enabled }}
{{ .Values.rdma.resourceName }}: {{ .Values.rdma.resourceCount | default 1 }}
{{- end }}
@ -199,6 +205,9 @@ spec:
nvidia.com/gpu: "{{ .Values.resources.gpuLimit }}"
memory: {{ .Values.resources.memoryLimit }}
ephemeral-storage: 10Gi
{{- if and .Values.resources.gpuMem (gt (int .Values.resources.gpuMem) 0) }}
nvidia.com/gpumem: {{ .Values.resources.gpuMem }}
{{- end }}
{{- if .Values.rdma.enabled }}
{{ .Values.rdma.resourceName }}: {{ .Values.rdma.resourceCount | default 1 }}
{{- end }}
@ -206,6 +215,9 @@ spec:
ephemeral-storage: 10Gi
cpu: {{ .Values.resources.cpuRequest }}
nvidia.com/gpu: "{{ .Values.resources.gpuLimit }}"
{{- if and .Values.resources.gpuMem (gt (int .Values.resources.gpuMem) 0) }}
nvidia.com/gpumem: {{ .Values.resources.gpuMem }}
{{- end }}
{{- if .Values.rdma.enabled }}
{{ .Values.rdma.resourceName }}: {{ .Values.rdma.resourceCount | default 1 }}
{{- end }}

6
vllm-serve/templates/single.yaml
View File

@ -92,12 +92,18 @@ spec:
resources:
limits:
nvidia.com/gpu: "{{ .Values.resources.gpuLimit }}"
{{- if and .Values.resources.gpuMem (gt (int .Values.resources.gpuMem) 0) }}
nvidia.com/gpumem: {{ .Values.resources.gpuMem }}
{{- end }}
memory: {{ .Values.resources.memoryLimit }}
ephemeral-storage: 10Gi
cpu: {{ .Values.resources.cpuRequest }}
requests:
ephemeral-storage: 10Gi
cpu: {{ .Values.resources.cpuRequest }}
{{- if and .Values.resources.gpuMem (gt (int .Values.resources.gpuMem) 0) }}
nvidia.com/gpumem: {{ .Values.resources.gpuMem }}
{{- end }}
ports:
- containerPort: 8000
name: http