No options appear under Config

[logangage03]

Description

I am attempting to stand up Weaviate and Verba using Run.AI job configuration. I have successfully done this using a Verba docker-compose.yml file, but was attempting to use the docker.io/semitechnologies/verba:latest option. I have no changed my Weaviate configuration. I am able to reach the Verba instance however the options under Config do not appear. Within Settings > Admin Panel this is the information under Weaviate Information:

• Connected to
• Deployement: Custom
• Version: (spinning)
• Nodes: (spinning)
• Collections: (spinning)

Installation

• pip install goldenverba
• pip install from source
• [x ] Docker installation (from docker.io/semitechnologies/verba:latest)

If you installed via pip, please specify the version:

Weaviate Deployment

• Local Deployment
• Docker Deployment
• [ x] Custom Deployment
• Cloud Deployment

Configuration

Reader:
Chunker:
Embedder:
Retriever:
Generator:

Steps to Reproduce

1. I am using Run.AI Job builder to stand up Instances
2. Weaviate
   AUTHENTICATION_ANONYMOUS_ACCESS_ENABLED = true
   CLUSTER_HOSTNAME = node1
   DEFAULT_VECTORIZER_MODULE = none
   ENABLE_API_BASED_MODULES = true
   PERSISTENCE_DATA_PATH = /var/lib/weaviate
   QUERY_DEFAULTS_LIMIT = 25
3. Verba
   DEFAULT_DEPLOYMENT = Custom
   OLLAMA_URL = http://olama:11434
   WEAVIATE_URL_VERBA = <weaviate_url>

Additional context

N/A

[thomashacker]

Thanks for the issue! Can you share more information about what RunAI is and how you're using it. Are you receiving any console outputs, any error message in the frontend dev console or on your docker console? Let me know, thanks! 🚀

[logangage03]

Hello,

Run:AI is a Kubernetes-based software platform that automates the allocation of compute power to users. It's designed to speed up data science initiatives by: pooling available resources, dynamically allocating resources based on need, maximizing accessible compute, reducing operational costs, and automating resource management and orchestration.

Run:AI's features include:

• Fair-share scheduling
• Simplified multi-GPU distributed training
• Visibility into workloads and resource utilization
• Control for cluster admin and ops teams
• Multiple queues on top of Kubernetes
• Configurable fairness policies
• Policy engine
• Open architecture
• Deep visibility into AI workloads
• Guaranteed quotas of GPU resources

No error messages on the frontend dev console:

Here is an image of the frontend where is says "No Config found"

Here is an image of the Event History from Run.AI

Here are the logs on the container:

INFO: Will watch for changes in these directories: ['/Verba']
WARNING: "workers" flag is ignored when reloading is enabled.
INFO: Uvicorn running on http://0.0.0.0:8000 (Press CTRL+C to quit)
INFO: Started reloader process [1] using WatchFiles
�[38;5;4mℹ Couldn't connect to Groq (https://api.groq.com/openai/v1/)�[0m
/usr/local/lib/python3.11/site-packages/websockets/legacy/init.py:6: DeprecationWarning: websockets.legacy is deprecated; see https://websockets.readthedocs.io/en/stable/howto/upgrade.html for upgrade instructions
warnings.warn( # deprecated in 14.0 - 2024-11-09
/usr/local/lib/python3.11/site-packages/uvicorn/protocols/websockets/websockets_impl.py:14: DeprecationWarning: websockets.server.WebSocketServerProtocol is deprecated
from websockets.server import WebSocketServerProtocol
INFO: Started server process [134]
INFO: Waiting for application startup.
INFO: Application startup complete.
�[38;5;4mℹ Couldn't connect to Groq (https://api.groq.com/openai/v1/)�[0m
�[38;5;4mℹ Cleaning Clients Cache�[0m
�[38;5;4mℹ Cleaned up 0 clients�[0m
�[38;5;4mℹ 0 clients connected�[0m
INFO: 10.42.0.218:57382 - "GET /api/health HTTP/1.1" 200 OK
�[38;5;4mℹ Cleaning Clients Cache�[0m
�[38;5;4mℹ Cleaned up 0 clients�[0m
�[38;5;4mℹ 0 clients connected�[0m
INFO: 10.42.0.218:57382 - "GET /api/health HTTP/1.1" 200 OK
�[38;5;4mℹ Cleaning Clients Cache�[0m
�[38;5;4mℹ Cleaned up 0 clients�[0m
�[38;5;4mℹ 0 clients connected�[0m
INFO: 10.42.0.218:57382 - "GET /api/health HTTP/1.1" 200 OK
INFO: 10.42.0.218:57382 - "POST /api/connect HTTP/1.1" 403 Forbidden
/usr/local/lib/python3.11/site-packages/websockets/legacy/server.py:1177: DeprecationWarning: remove second argument of ws_handler
warnings.warn("remove second argument of ws_handler", DeprecationWarning)
INFO: ('10.42.0.218', 57394) - "WebSocket /ws/generate_stream" [accepted]
INFO: connection open
INFO: ('10.42.0.218', 57406) - "WebSocket /ws/import_files" [accepted]
INFO: connection open
�[38;5;4mℹ Cleaning Clients Cache�[0m
�[38;5;4mℹ Cleaned up 0 clients�[0m
�[38;5;4mℹ 0 clients connected�[0m
INFO: 10.42.0.218:57382 - "GET /api/health HTTP/1.1" 200 OK
INFO: 10.42.0.218:57382 - "POST /api/get_meta HTTP/1.1" 403 Forbidden
INFO: 10.42.0.218:51794 - "GET /static/_next/static/media/dd4ab5b525bd804a-s.woff2 HTTP/1.1" 200 OK
�[38;5;4mℹ Cleaning Clients Cache�[0m
�[38;5;4mℹ Cleaned up 0 clients�[0m
�[38;5;4mℹ 0 clients connected�[0m
INFO: 10.42.0.218:33834 - "GET /api/health HTTP/1.1" 200 OK
INFO: 10.42.0.218:33834 - "POST /api/get_all_suggestions HTTP/1.1" 403 Forbidden
/usr/local/lib/python3.11/site-packages/websockets/exceptions.py:116: DeprecationWarning: ConnectionClosed.code is deprecated; use Protocol.close_code or ConnectionClosed.rcvd.code
warnings.warn( # deprecated in 13.1 - 2024-09-21
INFO: connection closed
INFO: connection closed
�[38;5;3m⚠ WebSocket connection closed by client.�[0m
�[38;5;3m⚠ Import WebSocket connection closed by client.�[0m
�[38;5;4mℹ Cleaning Clients Cache�[0m
�[38;5;4mℹ Cleaned up 0 clients�[0m
�[38;5;4mℹ 0 clients connected�[0m
INFO: 10.42.0.218:33834 - "GET /api/health HTTP/1.1" 200 OK
INFO: 10.42.0.218:33834 - "POST /api/get_all_suggestions HTTP/1.1" 403 Forbidden
�[38;5;4mℹ Cleaning Clients Cache�[0m
�[38;5;4mℹ Cleaned up 0 clients�[0m
�[38;5;4mℹ 0 clients connected�[0m
INFO: 10.42.0.218:47668 - "GET /api/health HTTP/1.1" 200 OK
�[38;5;4mℹ Cleaning Clients Cache�[0m
�[38;5;4mℹ Cleaned up 0 clients�[0m
�[38;5;4mℹ 0 clients connected�[0m
INFO: 10.42.0.218:47668 - "GET /api/health HTTP/1.1" 200 OK
�[38;5;4mℹ Cleaning Clients Cache�[0m
�[38;5;4mℹ Cleaned up 0 clients�[0m
�[38;5;4mℹ 0 clients connected�[0m
INFO: 10.42.0.218:47668 - "GET /api/health HTTP/1.1" 200 OK
INFO: 10.42.0.218:47668 - "POST /api/connect HTTP/1.1" 403 Forbidden
INFO: ('10.42.0.218', 47668) - "WebSocket /ws/generate_stream" [accepted]
INFO: connection open
INFO: ('10.42.0.218', 47676) - "WebSocket /ws/import_files" [accepted]
INFO: connection open
�[38;5;4mℹ Cleaning Clients Cache�[0m
�[38;5;4mℹ Cleaned up 0 clients�[0m
�[38;5;4mℹ 0 clients connected�[0m
INFO: 10.42.0.218:47684 - "GET /api/health HTTP/1.1" 200 OK
INFO: 10.42.0.218:47684 - "POST /api/get_meta HTTP/1.1" 403 Forbidden
�[38;5;4mℹ Cleaning Clients Cache�[0m
�[38;5;4mℹ Cleaned up 0 clients�[0m
�[38;5;4mℹ 0 clients connected�[0m
INFO: 10.42.0.218:51336 - "GET /api/health HTTP/1.1" 200 OK
INFO: 10.42.0.218:51336 - "POST /api/get_meta HTTP/1.1" 403 Forbidden
�[38;5;4mℹ Cleaning Clients Cache�[0m
�[38;5;4mℹ Cleaned up 0 clients�[0m
�[38;5;4mℹ 0 clients connected�[0m
INFO: 10.42.0.218:51336 - "GET /api/health HTTP/1.1" 200 OK
INFO: 10.42.0.218:51336 - "POST /api/get_all_documents HTTP/1.1" 403 Forbidden
�[38;5;4mℹ Cleaning Clients Cache�[0m
�[38;5;4mℹ Cleaned up 0 clients�[0m
�[38;5;4mℹ 0 clients connected�[0m
INFO: 10.42.0.218:51336 - "GET /api/health HTTP/1.1" 200 OK
INFO: 10.42.0.218:51336 - "POST /api/get_all_documents HTTP/1.1" 403 Forbidden

[thomashacker]

Thanks for the info! As I can see in the logs, you're getting 403 Forbidden when the frontend is interacting with the backend. This is due to the middleware that checks whether requests have the same origin as the backend. I think since you're running Verba on an external server, you're getting these errors.

[thomashacker]

Created an issue for it, will let you know once we fixed it! 🚀 #353