Arqum Younus

post-migration checks are where most moves go wrong: permalinks, mixed content after an https change, and scheduled tasks. (i’m with cloudways, so take this with salt) staging a copy first is a good habit, and it’s built into our platform. what’s the most common failure you see after a move?

a checklist like this is what stops a migration turning into panic at 2am. two things i’d add if they’re not covered: lower the dns ttl a day before the switch, and keep the old host live until email and forms are confirmed working on the new one. disclosure: i work at cloudways, where there’s a free wordpress migration plugin, but the order of steps is the same on any host. do you test the new server with a hosts-file entry before touching dns?

testing a restore on a copy instead of the live site is the part most people skip, and it’s the only way to know a backup is real. one thing worth checking on the copy: that the database and uploads both came back and you can actually log in, not just that the homepage loads. (cloudways team here, so biased) on-demand backups of an app’s files and database plus a staging clone are built in on our side, but the habit matters more than the tool. do you also time the restore so you know the worst-case recovery time?

the www vs non-www section and the old BASE_URL / NEXTAUTH_URL warning are the parts that save people the most time. two small things to add: a CAA record that doesn’t allow letsencrypt.org will block issuance, and wildcard certs need the DNS challenge instead of the HTTP one. (i’m on the cloudways team, where let’s encrypt install and renewal is built in, but these checks apply on any host.)

the time-cost step is the part most “self-host to save money” posts skip, and i like that you say the savings only hold if someone is willing to own the server. one thing i’d add for the client-site part of that list: if some of those are wordpress or php sites, there’s a middle option, a managed layer on top of a vps. you pay more than for a bare server, but backups, ssl and support are handled for you. (cloudways team here, so biased.) it only covers php apps like wordpress, laravel and magento, cloudways now also has managed node.js hosting (velocity), so node apps don’t have to stay on the vps. it’s git-based deploys with backups, ssl and patching handled, on flat monthly pricing, which ties into the cost math in your post. do you count backup restores in the monthly time estimate, or only deploys?