Self-Hosting Still Pays

You don't even need to buy a new EPYC or Xeon Scalable system. A 20 core Xeon v3/v4 system with 128gb of ram can be had off lease for $1000. Psychz.net provides 1u colocation with 1/Gbs link (10/Gbs burst) for $80/mo.

Equivalent EC2 is $300-600/month with no better reliability. Plus you have to deal with noisy neighbors and bandwidth overages. Build your baseline usage in colo and burst into AWS unless you like buying Bezos more houses.

I wondered whether this comparison (like many others like it) would /finally/ bake cluster operational costs into the calculus. Not this time, it looks like.

To raise a more constructive point. Looking at [1] (an earlier post in the series), it looks like their workload is WordPress + VBulletin forums, and "pets" not "cattle" -- I wonder how much more they'd be able to get from separating their stateful and stateless layers more cleanly, and using some of the more powerful (and cheaper) AWS primitives for serving traffic that involves the latter. Why do they need such beefy instances for essentially serving content?

The truth is, it's because the platforms they're using (vBulletin + WP) exist, they're very powerful and they get the job done. This leads me to believe that the operational UX for running these kinds of common applications (which compose most of the internet) on AWS without highly technical supervision is not great.

[1] https://www.servethehome.com/falling-sky-part-3-evaluating-a...

I see so often developers state something along the lines of, "We need to go with AWS!" (sometimes substituting AWS with GCP or Azure) and the reason they always give is "scalability". 99% of the time, what they're building not only doesn't need to scale to that degree but they somehow seem to think that it's always inherently cheaper.

Meanwhile, when I build something for a client my first go-to is something like DigitalOcean or Amazon LightSail. I know, LightSail is still AWS. But it's not going to automatically increase the bill by 300% because someone accidentally toggled something in AWS's byzantine admin console.

When ("if" is more often the case) that client starts to see the kinds of traffic that necessitates a full-scale AWS build out then I'll advise them of their options. By that point, they're actually generating real revenue and, as importantly, they didn't pay me to prematurely optimize a product that wasn't yet generating any revenue, which helped them succeed.

TIL: AWS is only expensive if your time has no value. Article neglected to add the cost of labour for setup (edit: aside of racking), maintenance (security updates on infrastructure, etc), and disaster response. It also missed the cost of downtime based on projected MTTR for a failure. The word "fail" only appears once in the page, and not in regards to recovery, only preventative replacement.

It's absolutely true that colo is cheaper than AWS for projects that can absorb the labour and failure related costs, but that usually is only true for hobby or very small business services.