Without docker and separate machines for front end and backend

[gl2007]

Just the title; very new to configuring celery, docker, redis, mongo to be setup for a separate machine (say W), and external LLM setup. Have Ollama setup on W with different models to try out.

I want to setup Frontend on my laptop, while backend, with celery, mongo, redis on W. I have setup redis on WSL. Wiring everything up without docker seems to be very involved process.

Can a guide be written to simplify such a setup? External LLM I have seen a guide so probably can manage that but everything else, I will probably spend 2-3 weeks to learn and get it working.

Any help will be greatly appreciated!!

[gl2007]

Got it working if anyone is interested. Here is my setup:
Laptop:
Windows and VS code
Frontend, backend & worker

Workstation (with optional GPU):
Windows
Ollama (various LLM options)
Memurai - Redis does not directly work in Windows, tried via WSL, big hassle to connect from host/or my laptop using intranet IP. ChatGPT suggested Memurai Redis alternative for Windows, so far working well
MongoDB

Here is my working .env:

API_URL=http://127.0.0.1:7091
API_KEY=OLLAMA_NO_KEY
#LLM choices are: openai,azure_openai,sagemaker,huggingface,llama.cpp,anthropic,docsgpt,premai,groq,google
LLM_NAME=openai
#LLM_NAME=docsgpt
#below is used if LLM is hosted by "backend"
#MODEL_NAME=svk/docsgpt-7b
MODEL_NAME=docsgpt-7b-f16
MODEL_PATH=models/docsgpt-7b-f16.gguf

EMBEDDINGS_NAME=huggingface_sentence-transformers/all-mpnet-base-v2
EMBEDDINGS_KEY=OLLAMA_NO_KEY

#Ollama OpenAI compatible API
OPENAI_BASE_URL=http://192.168.1.xyz:11434/v1/
#llama-cpp-python OpenAI compatible API
#OPENAI_BASE_URL=http://192.168.1.xyz:8000/v1/

MONGO_URI=mongodb://192.168.1.214:27017/docsgpt
CELERY_BROKER_URL=redis://192.168.1.xyz:6379/0
CELERY_RESULT_BACKEND=redis://192.168.1.xyz:6379/1
CACHE_REDIS_URL=redis://192.168.1.xyz:6379/2