Add host override to query specific pod

[varungup90]

Query specific pod by including pod's ip address and port

curl -v http://localhost:8888/v1/chat/completions \
  -H "target-pod: 10.244.1.3:8000" \
  -H "user: varun" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer any_key" \
  -d '{
     "model": "llama2-70b",
     "messages": [{"role": "user", "content": "Say this is a test!"}],
     "temperature": 0.7
   }'

[Jeffwan]

overall looks good to me

[Jeffwan]

I may miss some of the contexts, I think our goal is to implement the routing part. Why do we care the target ip support?

[varungup90]

I may miss some of the contexts, I think our goal is to implement the routing part. Why do we care the target ip support?

Routing algorithm will select a pod, and target-ip is to forward request to that pod.

[Jeffwan]

/lgtm