second checkpoint

This commit is contained in:
Kai Wu 2025-08-02 13:16:35 -07:00
parent 67f19f76b2
commit 645e55a450
4 changed files with 78 additions and 17 deletions

View file

@ -26,12 +26,34 @@ spec:
app.kubernetes.io/component: server
spec:
initContainers:
- name: wait-for-vllm-server
- name: wait-for-services
image: busybox:1.28
command: ['sh', '-c', 'until nc -z vllm-server.default.svc.cluster.local 8001; do echo waiting for vllm-server on port 8001; sleep 2; done;']
- name: wait-for-llm-nim-code
image: busybox:1.28
command: ['sh', '-c', 'until nc -z llm-nim-code.default.svc.cluster.local 8000; do echo waiting for llm-nim-code on port 8000; sleep 2; done;']
command: ['sh', '-c', '
echo "Waiting for all required services to be ready...";
echo "Checking vllm-server...";
until nc -z vllm-server.default.svc.cluster.local 8001; do
echo "waiting for vllm-server on port 8001";
sleep 2;
done;
echo "vllm-server is ready!";
echo "Checking llm-nim-code...";
until nc -z llm-nim-code.default.svc.cluster.local 8000; do
echo "waiting for llm-nim-code on port 8000";
sleep 2;
done;
echo "llm-nim-code is ready!";
echo "Checking ollama-safety...";
until nc -z ollama-safety.default.svc.cluster.local 11434; do
echo "waiting for ollama-safety on port 11434";
sleep 2;
done;
echo "ollama-safety is ready!";
echo "All services are ready!";
']
containers:
- name: llama-stack
image: llamastack/distribution-starter:0.2.15
@ -56,7 +78,7 @@ spec:
- name: NVIDIA_BASE_URL
value: http://llm-nim-code.default.svc.cluster.local:8000
- name: OLLAMA_BASE_URL
value: http://ollama-safety.default.svc.cluster.local:8000
value: http://ollama-safety.default.svc.cluster.local:11434
- name: POSTGRES_HOST
value: postgres-server.default.svc.cluster.local
- name: POSTGRES_PORT
@ -69,7 +91,7 @@ spec:
value: "${CODE_MODEL}"
- name: TAVILY_SEARCH_API_KEY
value: "${TAVILY_SEARCH_API_KEY}"
- name: OLLAMA_MODLE
- name: OLLAMA_MODEL
value: "${OLLAMA_MODEL}"
command: ["/bin/sh"]
args: