[BUG]: White screen once accessing llm-preference

[ryan-damen-rz]

How are you running AnythingLLM?

Not listed

What happened?

Hello, I am using the anythingLLM render docker image and deploying it through k8s. Whenever I enter the settings menu and navigate to AI Providers -> LLM the web page turns white. Opening any other setting window goes fine without any issue.
In a separate workspace I have an 22 days old anythingLLM render docker image that does not have this issue.

Are there known steps to reproduce?

1. Use mintplexlabs/anythingllm:render as image for your k8s deployment.
2. Navigate to the webpage (chrome or edge)
3. Navigate to Settings -> AI Providers -> LLM
4. Screen turns white

[timothycarambat]

You dont have the ENV file set up correctly - which is why you are getting that crash.

Also why are you using the render image on K8s? That docker image is for Render.com/Railway.app - otherwise you should be using latest

Also there is a k8 manifest
https://github.com/Mintplex-Labs/anything-llm/blob/master/cloud-deployments/k8/manifest.yaml

[ryan-damen-rz]

The initial deployment I had is based of the k8s manifest you shared and it used to work.

I think the decision to use the render image is correct as that manifest also uses the render image:
image: "mintplexlabs/anythingllm:render" as such I did the same.

To which ENV file are you referring to? The only ENV variables I can find in the manifest have been set up identically in my deployment are there any other ENV files that need to be set up?

[timothycarambat]

You did not add anything to this section?

- name: AWS_REGION
            value: "{{ aws_region }}"
          - name: AWS_ACCESS_KEY_ID
            value: "{{ aws_access_id }}"
          - name: AWS_SECRET_ACCESS_KEY
            value: "{{ aws_access_secret }}"
          - name: SERVER_PORT
            value: "3001"
          - name: JWT_SECRET
            value: "my-random-string-for-seeding" # Please generate random string at least 12 chars long.
          - name: STORAGE_DIR
            value: "/storage"
          - name: NODE_ENV
            value: "production"
          - name: UID
            value: "1000"
          - name: GID
            value: "1000"

If so, what should happen is on deployment is that there should be an .env file in this mount

volumeMounts: 
         - name: anything-llm-server-storage-volume-mount
           mountPath: /storage            

That should have all the current settings for the ENV of the system on startup. I dont know what your container logs look like as that would provide more context here. These templates are community created so I cannot assure they work - I know some people had success with them so they are published for that reason.

[ryan-damen-rz]

I have solved the issue!
Originally I was planning to enter the LLM through the settings but this generated a white screen. So instead I configured the LLM through environmental variables. This has solved the issue as the white screen does not appear anymore and I am freely able to change the LLM afterwards through the GUI. Could it be that there is no default LLM set in the render version causing the error?