Show Notes
Every time you load a website, send an email, or update an app, you're quietly relying on a handful of unglamorous services that route your packets to the right place: DNS to translate names into addresses, and BGP to figure out how to actually get there. When these systems break, or get attacked, the Internet doesn't just slow down but stops working.
For more than 25 years, NLnet Labs has been one of the small, non-profit teams keeping that core infrastructure running. Their software, including the DNS servers NSD and Unbound, the RPKI tools Krill and Routinator, and the new DNSSEC signer Cascade, is deployed everywhere from hobbyist Pi-Hole setups to Let's Encrypt and major Internet operators. And increasingly, it's written in Rust!
In this episode, I talk to Arya Khanna and Martin Hoffmann from NLnet Labs about what it takes to maintain critical Internet infrastructure as a small team, why they bet on Rust for new projects like the domain crate and Cascade and what the rest of us can learn from a codebase whose users include the people who keep your routes flowing.
More from Rust in Production
Welcome to the final episode of this season of Rust in Production. My guest is Orhun Parmaksız from JetBrains, and we talk about building developer tools with Rust.
Most Rust developers use the language, compiler, package registry, and tooling every day without thinking too much about the organization that helps keep parts of that ecosystem funded and sustainable. This episode is a re-introduction to the Rust Foundation: what it does, what it does not do, how it relates to the Rust Project, and why that distinction matters for teams using Rust professionally.
Most Rust in Production stories are about scale and performance. This one is a story about low-cost phones and patchy mobile connections in Africa, where a student is learning maths over WhatsApp. The whole point is to support hundreds of thousands of students cheaply enough to run at government scale.
There's a particular kind of pressure that comes with maintaining software at the very bottom of someone else's stack. ClickHouse lives in exactly that spot: roughly 1.5 million lines of mostly C++ and tens of millions of tests every single day.
I don't know about you, but to me there are few things as interesting as the hardware/software interface: the point where carefully written code meets the messy, physical world of sensors, lenses, and real-time constraints. It's where a clever abstraction either holds up or falls apart the moment a real signal hits it.