Notes from the field

Practical writing on Linux, virtualization, hosting, cloud, databases, security and keeping systems running, one post per service we offer.

Managed vs. Self-Managed Servers: What You're Actually Paying For

"Managed" and "self-managed" get compared on price, but the real difference is who's accountable when something breaks.

Turning off password auth is where most SSH hardening guides stop. Here's what actually closes the gap.

Docker doesn't remove operational work, it moves it. Here's where container setups tend to quietly fall apart, and how to keep that from happening.

24/7 coverage isn't free, and not every business needs it. Here's a straightforward way to decide.

Migrations go wrong when they're rushed or undocumented. Here's the staged approach that keeps a move from becoming an outage.

The metrics that predict an outage are usually already there. The problem is nobody's watching them, or the alerts are too noisy to trust.

3 copies, 2 media types, 1 off-site. Simple to say, and most backup setups still get at least one part of it wrong.

Most server compromises aren't sophisticated. They exploit the same handful of unhardened defaults. Here's the checklist that closes them.

Both databases are good enough for almost everything. Here's what should actually drive the decision instead of habit or hearsay.

Nobody notices their AWS bill go up 4% a month until it's doubled in a year. Here's where that growth usually hides.

The Nginx vs Apache debate is mostly settled by what you're running, not which one is 'better.' Here's the practical version of that decision.

Virtualization isn't free. Here's how to tell whether splitting a server into KVM virtual machines will actually solve a problem you have.

Patching gets skipped because it feels risky. Here is a cadence that keeps a server current without turning every update into a gamble.