Three sites moved to new infrastructure with no maintenance page, no downtime, and nothing a visitor could feel.
We moved three sites to new infrastructure, and nobody visiting them could tell anything happened. No maintenance page. No “we’ll be right back.” No broken checkout for a few hours at 2am. That’s the part we’re proud of, and it’s the part nobody sees, which is the whole point.
Let’s be honest about what “zero downtime” actually is, because it isn’t luck and it isn’t magic. It’s mostly sequencing and patience. When you point a domain at a new server, the internet doesn’t update all at once. Every network holds onto the old address for a while, based on the TTL, the time-to-live on the DNS record. Think of DNS as the post office’s forwarding card for your domain. Drop a new card the day you move and half your mail still goes to the old house for a day or two while everyone catches up.
So you don’t do it the day you move. A week ahead, you lower that TTL way down, so when you finally flip the address the whole internet catches up in minutes instead of days. Then you build the new site fully on staging, copy the database over, and check it against the live site until it’s boring. Boring is the goal. The actual cutover, the moment that scares everyone, is the smallest step of the whole job, because by then you’ve already proven the new place works.
Here’s why this matters to you and not just to us. Downtime during a migration isn’t a tech inconvenience, it’s a business one. It’s the form that won’t submit while you’re running an ad straight at it. It’s the support tickets. It’s the marketing lead who has to email her CEO and explain why the site was down. no downtime, no lost pages, and no support tickets during the cutover We ran all three of these the same careful way, and the best outcome was that it was unremarkable.
We’d rather a migration be the most boring week of your quarter than the most memorable one.