So, for the API, the machine health should be indeed provided. From the UI:
1. when tests other than cpu/ram/memory have failed, have the "overall health status" icon should be shown (with proper tooltip). This would fix https://bugs.launchpad.net/maas/+bug/1721823
2. when tests like cpu/ram/memory have failed, the "overall health status" should be *blank* (the same as how it is today).
3. when user has overriden the "failed testing" state (from commissioning/testing/ready), show the "overall health status" icon (with proper tooltip e.g. "tests have failed, but status was overridden" alongside all other icons)
4. when the machine was deployed -> testing -> deployed -> if there are failed tests, it goes back to deployed with the "overall health status" icon
So, for the API, the machine health should be indeed provided. From the UI:
1. when tests other than cpu/ram/memory have failed, have the "overall health status" icon should be shown (with proper tooltip). This would fix https:/ /bugs.launchpad .net/maas/ +bug/1721823
2. when tests like cpu/ram/memory have failed, the "overall health status" should be *blank* (the same as how it is today).
3. when user has overriden the "failed testing" state (from commissioning/ testing/ ready), show the "overall health status" icon (with proper tooltip e.g. "tests have failed, but status was overridden" alongside all other icons)
4. when the machine was deployed -> testing -> deployed -> if there are failed tests, it goes back to deployed with the "overall health status" icon