Access milvus vector DB service within VSS blueprint

I have launchpad access with 8 NVIDIA H100 NVL GPUs. I am able to deploy and use VSS with nvila successfully, but I am facing issue when trying to access milvus service within VSS blueprint. It is showing it’s running on port 30848 and 32454 but when I am trying to use it using <NODE_IP>:PORT then I am getting 404 error. I have also attached the logs from milvus service.

These are the commands and link from the documentation I followed:
( VSS Customization — Video Search and Summarization Agent)
milvus_logs.txt (4.7 MB)

kubectl patch svc milvus-milvus-deployment-milvus-service -p ‘{“spec”:{“type”:“NodePort”}}’
kubectl get svc milvus-milvus-deployment-milvus-service # Get the Nodeport

Could you attach the specific commands you used?

kubectl patch svc milvus-milvus-deployment-milvus-service -p ‘{“spec”:{“type”:“NodePort”}}’

Did the VSS report a 404 error when you were running this command?

No, VSS UI is working fine as expected but not able to access milvus through the above port.

Could you attach the specific method you access the Milvus on your side?

<NODE_IP>:PORT
Where NODE_IP is same what I used for VSS UI and able to access it and tried PORT 30848 and 32454 but it’s showing 404 error.

Milvus is a database. You cannot check that just from the web. You can learn some basic concepts about Milvus and check that with command or python code.

What this documentation is referring to?


This is just teaching you how to obtain the ip and port. After obtaining them, please learn how to use the milvus_cli or Milvus Python SDK to access the database.

This is the services running and ports on which they are running:


I tried accessing milvus using CLI but still not able to do so getting below errors:

In your scenario, it should be using port 31909. We need to investigate why does it report an list index out of range error after the link is successfully established.