Why I’m Starting a Homelab as an Aspiring Data Analyst

A good way to think about a homelab (IC)

To say that trying to break into a data-based career is overwhelming is a massive understatement. The vast amount of acronyms, technologies, and skills required feels like diving into an ocean at night without a headlamp. Although there are a ton of helpful (and high quality) tutorials online, it can feel impossible to know where to begin.

Combine the confusion of where to begin, with the fear of breaking something in production at my current job, and I was feeling a bit hopeless.

I want to become one of the best data analysts around. I love stories and mathematics (as evidenced by my writing and reading here.) Data analytics feels like the perfect fusion of the “left” and “right” sides of my brain.

To me, data analytics is like being a scout — I wade into the wilderness of data, take or jot down what is needed, and then pop back out. Each request is an adventure and my skills are my tools to get the job done.

I’ve also fallen in love with model creation. I have a love for science that started when I was young and has never gone away. Great scientists were and still are my heroes. I feel like a bit of a scientist myself when I make models — it’s an awesome feeling.

Why?

So, enough about me and my love for data and science. Why would someone who has similar aspirations as me take on homelabbing?

As I said earlier, being a data analyst is like being a scout, constantly wading into the unknown with only your wits and tools to keep you alive. Just like our favorite heroes need to constantly train their skills to stay sharp and experiment with new techniques, so too does a data analyst.

When you’re working for a nonprofit and they’re paying you for a result, they are rarely interested in your development beyond the point of their needed returns. The corporate environment is not a safe place to hone your craft (in broad terms.) Instead, when you’re working for someone else, they need you to get what they need in the most efficient manner possible.

So, if my work cannot be the playground and training ground I desire, I need another solution.

To the internet! The start of all good rabbit holes! (IC)

Through a lot (and I mean a lot) of desperate Googling, I finally stumbled upon the idea of a homelab. Admittedly, most of the contexts I saw it in were for IT professionals. The idea being a place you can learn about how to manage switches, connections, servers, etc.

Thankfully, the motto of data analysts is to adapt and overcome. I immediately saw the ability to repurpose the same idea for my needs (and experiment with cool additional techy stuff on the side.) I realized if I could learn a bit about hosting a server then I would have a safe place to store large(er) databases and practice my skills.

At last! I found a way to affordably create my mini production environment to learn about things like the cloud, large SQL databases, and to host my (probably overambitious) projects. As a bonus, I could flesh out my skillset as more of a “full-stack” data analyst.

Of course, everything has its price to entry and homelabbing was no exception.

How I felt at first trying to understand this new hobby… (IC)

Once again, I found myself feeling overwhelmed and confused with another new world of information. Thankfully, unlike before, I’m a little bit more seasoned and knew what I wanted. Harnessing the power of Google-Fu, I slowly found my way to some beginner resources.

I also remembered two quotes that helped me push through the initial confusion and frustration of trying to figure this out, “just do it” and “perfect is the enemy of good.” No doubt, there are a million ways to accomplish what I want better (and I’ll be asking on one of the forums I suggest.) I’m not too concerned about the mistakes though — they’re part of the learning and I’d rather be moving in a direction and correct course than never start at all.

Below are some of the resources that really helped me on my journey and I leave them here in the hopes that someone wanting to do something similar can be saved some of the time I took.

Resources for Learning:

Forums:

https://reddit.com/r/datahoarder

https://www.reddit.com/r/homelab/

https://www.reddit.com/r/selfhosted/

YouTubers:

https://www.youtube.com/user/NetworkChuck

https://www.youtube.com/c/TechnoTimLive/videos

Goals and Current Setup

So, after my research and enthusiasm, what goals did I set and what setup did I arrive to?

Goals:

• Server to host large datasets
• Use these datasets for portfolio projects, and to practice more in-depth machine learning
• File storage (linked to an off-site backup) for the following items:
• Important Personal Documents
• STL and Gerber Files (Hobby electronics is going to happen soon)
• Woodworking Plans (I want to be a maker, okay?)
• eBook, eCourse, and webcomic library (I’ve got a small collection of eBooks/eCourses and want to store my favorite webcomics locally.)
• Webserver (In the future, will buy a PI to do this)
• Pi-Hole! (Future)
• Learn Docker, Kubernetes, and AWS (Future)
• Maybe self-host through Owncloud
• Media Server (but in a non-traditional way, stay tuned!)

With these goals in mind, I present, my humble homelab V1:

• TerraMaster F2–220 NAS with 2, 2TB drives in a RAID 1 configuration (Yay eBay sale!)
• Dell Precision T5400 (Was previously used and again, eBay sale)
• Old laptop (for a media server, if desired)

One thing I learned after the fact — check for power draw!

Turns out, not only is the T5400 super heavy but it is also an energy guzzler. For my current use case, this isn’t as much of an issue. I don’t need 24/7 uptime since I’m not using the server as a media server. I have an old laptop that I could connect and use as a media server if needed, which is my short-term plan to meet that need.

In the near future, I plan to set up my NAS and start moving some files to it, establish an offsite backup, and find a home for my homelab (photos soon)!

Conclusion

So, in my quest to become a better data analyst, I opened up three new hobbies (homelabbing, self-hosting, and hobby electronics/3d printing) that I’m excited to dive into!

Although homelabs have traditionally been used in IT, I believe data professionals can benefit from the practice too. The creation of a custom playground to learn, play, and make mistakes judgment-free is invaluable.

In the coming posts, I plan on detailing my (mis)adventures in setting up my homelab for data analytics and other quality of life projects. I hope you decide to join me on this journey and that this served as some inspiration.

I also hope this is the start of a conversation. If you have thoughts, advice, tips, etc, please drop a comment down below!

For now, venture bravely into the wilderness, adventure is out there!