We just shipped a great improvement to your app machines dashboard.
Per process group separation
We know process groups are meant to do different things in your app so it makes total sense to separate the machines list per process group so you can easily reason about your current scale.
Plan scaling machines easily
You used to need to run a series of fly scale [vm/count/memory] to get a desired final scale for your app. This ends today, you can do everything in a single UI now:
Plan your changes by machine size + count per region + memory MB then hit Verify Changes to see what’s gonna happen.
Only after hitting Confirm you will actually apply changes to your machines.
App overview now directs you to this page
We are keen in telling customers to run apps in more than one machine and more than one region because we believe in the power of high availability. Rather than giving a CLI command now we will send you to the machines page to scale your app machines.
Out of scope for now
Currently we scoped out some cases:
- GPU Machines and Machines with volumes: we will still tell you about
fly machine clonecommands for now. - Unmanaged postgres: we don’t plan on adding an specific scale machines UX for this kind of app in the near future but we believe that by supporting machines with volumes this will be also covered
- If your app has machines with different images this could lead to unexpected behavior so we will warn you upfront about this but we wont block you from using this feature.
Let us know what you think about this one! Is your use-case different from what was covered here? Let us know by dropping a comment