Hello everyone. This video has been a

long time in coming. I wanted to do a

quick run through of my home lab. As you

can see here, everything is displayed on

homepage. And even though my hardware is

fairly modest, I am able to run quite a

bit without using too much of the

available resources. So recently I

upgraded from a Raspberry Pi 4 with 8

gigs of RAM to an orange Pi 6 Plus which

you can see here. I will include all the

links to the GitHub for my home lab as

well as the various hardware that I'm

using. So when you look here, this is

actually a very new piece of hardware. I

believe it was October or November of

2025 it was released, but it is running

very well uh with my whole Kubernetes

stack. I'm using Iuntu 26 on that and I

will have a video going more in detail

on that and the experience so far. The

other piece of hardware that I have is

an HP ProBook 430 G6 which you can see

you can get for a fairly reasonable

price used on eBay which is what I did.

I think it cost me about 120 bucks and

it has 100 sorry 16 gigs of RAM and an

i3 processor.

So yeah, in here on my homepage, you can

see I have a lot that I'm testing out

with regarding AI and LLMs. Open web UI

is my main chat that I use to test

various MCP servers that I'm working on

as well as different models and testing

out some free offerings from different

providers. I also have Llama, Quadrant

as my vector database, fresh RSS for

some RSS feeds, Libra chat as an

alternative to open web UI. Also, I'm

experimenting with Jupyter Notebooks and

also Phoenix. This is Arise Phoenix,

which allows you to hook up things like

Open Web UI, and you can actually go in

here and see some of the traces and

spans and different things that you have

uh to actually track the conversations

that you're having with your LLMs in

Open Web UI. I have some pretty standard

uh monitoring and observability setup in

Graphana and Prometheus as well. I can

see my whole home lab. I can see the

database size for my Postgress database

or database whatever I have as

well as uh some MCP server monitoring

here which does also show up uh when you

configure it in Arise Phoenix as well.

Here is my quadrant deployment still a

work in progress I'm hoping to make a

video on that soon. PG admin which just

helps me administer my Postgress

database context forge which is an MCP

gateway more or less with some

authentication

uh capabilities as well as some virtual

server capabilities where I can specify

the specific tools I want to expose from

the MCP servers that I'm running. I also

have MCPO uh which I believe is MCP

open it's an open- source proxy for

various MCP servers. So I use that for

some out of the box MCP servers like

Kubernetes and others that I'm working

with. I also have a local deployment of

AWX which is a front-end open source for

Anible. I'm doing some testing there. I

do also have N8N Jenkins Loki Coder some

various other things in my home lab that

I'm using to test out. I have Reddus as

my uh quick caching software to help u

with some of the performance of open web

UI and then also the performance of

context forge and then regarding how

this is all deployed. So it's a full

Kubernetes stack [clears throat] managed

through flux. So basically everything

that I have in my code which I can pull

up here.

All of this is operating under a GitOps

flow where it uses Flux to sync all the

manifests and the Helm charts and

everything else that I have in my

Kubernetes home lab which you can see

here. I'm not going to go into too much

depth today on how each of these is

configured, but if I just pull one up,

I'm using customize and it basically

says these are all the YAML configs that

I want. Flux will sync that to my

cluster and then deploy the pods

accordingly. So you can see all the pods

are running here. Uh if you're familiar

with Flux, you know generally how it's

working. But if I run this command, it's

going to pull down everything from my

GitHub repo and then sync it to my

cluster. Whatever is in that code, it

gets pushed as the config for this

cluster. So if I remove some config,

it's going to tear down those pods. Uh

and vice versa, if I add some new code,

it's going to spin it up from there. So,

it's very functional, especially for a

home lab. I only have, you know, one

worker node and one control plane node.

So, it's nothing crazy, just for my own

testing, but it does work very well. So,

that being said, and you can see the

homepage. Uh, I just refreshed because I

ran that Flux uh sync. But, yeah, let me

know if this was helpful. Uh, let me

know if you have questions about any of

these particular items or if you want me

to make a more in-depth video on any of

these pieces of my home lab. I do have

plans for quite a bit more regarding

MCP. I'll be releasing some videos uh

related to MCP servers that I have

running as well as NAN that hooks into

those MCP servers as well as various LLM

backends. And yeah, as always, I

appreciate you guys watching. Thanks.