EntwicklerHeld

We first needed to find a provider, a cloud provider. We had been at quite a number of them. We really had tried out everything. And the challenge with our infrastructure was to make it, well, migrate-able — because we didn't want to get locked into anything at the beginning. I think one of the biggest challenges is getting the VMs started up — for example — like when we need to scale up a lot and quickly. So that meant we would do a lot even before the provisioning stage, meaning, like starting up a lot of machines so that users could access their environments relatively quickly. That's the big advantage of Hetzner Cloud; we wanted to try them out — see if maybe we could even get the VMs to start dynamically — because it is just so quick.

Well, I would say the main reason is the low cost. It means we pay only one-tenth to one-twelfth the cost compared to other cloud providers that we had tried out and for the same configuration — but the support — and simply the fact that the cloud actually runs in Germany and with a German company — was a plus, too.

The migration, too, actually went really smoothly. And it's got modules for all the tools we need — Terraform, Ansible, and so on. So that made it go really quickly. We did the migration for all of our productive data. It took about two man-weeks to migrate our 40,000 users.

It triggered some new thoughts about how to do things — how we could make better use of the speed, for example — to improve our product even more and to give our users more opportunities, like, for example, that the VMs just start so quickly. So I don't need to wait for five minutes if I have bad luck, but instead, really just between five and ten seconds, and then the VMs are there and can handle the dynamic workload from our users. And whenever someone develops something on the platform, it runs on Hetzner Cloud so it's dynamically scaled and so that really all of the computational resources are available and are used to tackle the task. That means we can really, actually scale in a totally, totally dynamic way — so for each user, start up something here, and end something else there — and use that to still have great cost effeciency. That's what we've got planned. And it means, this inspiration that comes together with these improvements, and with getting more effiencient, it's good for the environment, in a way, because we're simply not using as many resources.

Naturally, we check on that — and we see that resources are saved. Like, maybe, How are the server rooms cooled? Or, how are resources saved by using Arm processors? Actually because we have a whole — because our load can have some big peaks and it can change per user. And, well, an Arm server has lots and lots of cores. And that's really suitable for this WebWelt ("WebWorld") — for this world, that we use.

We want to make EntwicklerHeld into a platform for developers that will be relevant, even in the future. I mean, not just for people who enjoy solving challenges, but also for people who want to consume — read something — a blog article here, other interesting content there — and to form a real community. And we want to do that together with companies where we say, "Okay, this is a connection that isn't gonna just disappear." And when it comes to other companies, our goal is actually to put companies in the spotlight that many developers just don't know about. and to draw out some "hidden champions", and to bring some attention to them via coding challenges on our platform. That is actually the vision that has been our driving force — right — and, of course, we want to keep growing in terms of user numbers. And we want to remain as dynamic as possible — agile — that's precisely our goal — our vision. We want to keep it fun!