Fusion powered web server
We already have fusion power. It comes from the sun. We just need to collect it.
The goal#
I have a new pet project. I want to build a solar powered web server. It occurred to me that I want to do this when I came across a post by Drupal founder Dries Buytaert who had done just this. The article also reminded me of an older post by Lowtech Magazine that I had read with great interest at the time.
I love the idea. I’m not really a techie, but I’ve been messing around with self-hosting for a while. I also know my way around the more obvious bits of a Linux system. That should be enough. (It should be said that I also know enough about self-hosting, and about my skillset, that all my production systems – email etc. – are hosted by professionals, whom I pay for their services.) I know very little about solar, except that it’s the future, and that I should learn to understand it a bit better.
I want this to be fun. A no-pressure learning experience. Some satisfying fiddling.
Requirements#
Functional#
If things go well, at the end of this I’m going to have a small web server, powered by solar energy, running a static website, reachable from the internet. The website in question would probably be this blog. (Bonus goal: host 1-2 other services too, maybe an RSS aggregator.)
I also want to document the process here. For others to learn and do the same thing, sure. But more importantly for myself, because this thing will run for a few months, and then something will change, and I’ll have forgotten everything about how it works.
Non-functional#
- The whole solution should be as simple and cheap as possible. If I have to choose between the two, then simple beats cheap - my time is valuable.
- It should be easy for others to replicate the solution. I want something barebones, just functional. Because realistically, I won’t be able to manage much more anyway.
- As a soft goal, I’d like to achieve an uptime north of 90%. That’s less than 2.5h of downtime per day.
How much downtime? Are you mad?#
No. I have a realistic appreciation of how important my personal blog is to humanity at large. If we’re prepared to accept a few rounding errors, then “not at all” is a fair answer. More philosophically: If you wanted to talk to me IRL, you’d pick a reasonable time of day. You wouldn’t expect me to get up for you at 3am in most cases. If you can handle this, you can handle my website being down.
OK fine. Now what?#
In the next post, I’ll draft some features and user stories for this whole enterprise. Let’s go for a ride!