Valérian de Thézan de Gaussan · Data Engineering for process-heavy organizations

Inexpensive Data Platform

Don't let yourself be fooled by crazy engineers.

☢ Before deploying heavy nuclear options for your data infrastructure, did you know a simple small machine might be sufficient?

You have pipelines to run, whether they are data integration pipelines between your systems, or pipelines to feed analytics, there’s no need for huge servers.

You can manage tens of millions of messages on a small machine like one with 4 cores, 8GB of RAM. That’s equivalent to a Raspberry Pi 4 for less than €100 💶. (I checked the prices this morning)

The vast majority of businesses don’t need hyper-complex solutions. A small machine with just 99.5% uptime often suffices. A small business can’t afford €1,000 per month in infrastructure just for a few pipelines with 10,000 messages/day (I’ve seen it!).

For example, if you’re with AWS, a small EC2 with 99.99% uptime will do: install your pipelines on it, run, and you’re done. Cost: less than 80€ per month, depending on the actual machine you choose.

👉 Upgrade only if and when the business has new requirements to maintain.