Zero downtime deployment is now the default on orkestr

I pushed broken migrations to production too many times.

Then, you know, the app crashed on startup, every request returned 502, and by the time I reverted the commit and redeployed, we had 17 minutes of downtime. Or worse, I had to do stuff on the production database. Who does that? Anyway, customers noticed the downtime. I still think about it at 3am.

That won’t happen on orkestr anymore.

How could we live without this until now? Honestly, I don’t know.

Starting today, every deploy uses a blue-green strategy by default - and it’s included on all plans, Starter through Team. Here’s what changed: when you push, we run the new version alongside the old one. The old version keeps serving traffic until the new one passes both a container health check and an application-level smoke test. Only then does traffic cut over. If you ship a bad deploy, your users never see it. The previous working version keeps serving traffic while you fix things and redeploy.

What could go wrong? (not much, honestly)

We thought about edge cases:

• Database migrations: Two versions of your app running at once is usually fine. If you have a destructive migration that the old code can’t handle, you already knew not to deploy that without downtime.
• Filesystem writes: If your app writes to the filesystem, both versions can write for a short time during the swap. The fix is the same: use a proper database, not filesystem writes, for anything that needs consistency.

In practice: deploy 50 times, and the worst thing that happens is you notice nothing. Your users notice nothing.

Why this matters

Every hosting platform talks about “zero downtime deploys.” Most of them mean “we’ll try really hard to not have downtime.” Vercel does this well. Railway does too. Heroku has been doing it for a decade.

But here’s the thing: orkestr gives you the same deployment guarantees as the big players, without the opaque pricing that explodes when you add a background worker. Pro is 9.99 EUR/month. Team is 29.99 EUR/month. No surprises.

And you don’t have to configure anything. It just works. Try it yourself - the first project is free.

What else is on the menu?

Blue-green is just the start. On the list: canary deploys (test with 5% of users), rolling updates (slow and steady), shadow mode (mirror traffic to test). Blue-green felt like the right default - simple mental model, zero downtime, no surprises. The others are coming.

Try it

Here’s to all the downtime we’ll never have again.

Deploy with confidence. Or reckless abandon. Either works now.

The feature is rolling out as the default for all new deploys. Existing projects get it automatically. Nothing to enable, nothing to pay extra for.

It’s the deploy strategy we wish we’d had that one night. Now you have it too.