Ok… maybe this is crazy, but the insanity over a ST failure/cloud failure whatever failure needs to be met with realistic expectations. The people of ST don’t wake up each morning bright eyed hoping things break. That being said failures are a reality. 100% of us (I am sure) have backups of data. Some of us, in business, use redundant internet connection and circuits. Why? Because we know something will fail even if that something is run by a fortune 500 company.
My rant is over… but what if the following.
I am not a programmer (now hurl insults), but could a ST app be created that effectively neutralized a cloud failure by simply shutting things down? That’s the challenge but what I am saying is that my automations stopped, my alarms stopped and everything went back to the way it was in 2010.
Last night, my lights started malfunctioning and other things that were annoyances. They were not as bad as some who had issues with SHM and alarms. But what if a mechanism existed (short of unplugging the hub) that would allow a user to simply turn off the smart home with a push of a button or an app or a CoRE.
Unfortunately there’s no smartapp that can run when smartapps aren’t running.
You can build a fully redundant second system as a Kill Switch pretty easily; it’s actually the same method that people use to make sure they can reboot the hub remotely if they need to. Just don’t use batteries in the hub and plug it into a smart plug which is controlled by a system other than SmartThings. (A WeMo plug works fine for this, but there are lots of other choices as well.)
Then when there’s a problem, you just use the other app to shut down your hub and bingo! Instant 2010 if that’s what you want.
As long as your Internet is still working, you could even do this remotely if the problems were occurring and you were not at the same location.
You just can’t do it with software within a software system that’s malfunctioning, because there’s no guarantee that the kill switch would be unaffected by the problems affecting everything else. But once you step outside that system, it’s pretty simple.
There are some ways you could do that, but you have to be able to define “goes crazy” in a way that you could capture as a rule. For the kind of problems that we saw yesterday, I think they unfortunately they were just too random to be able to do that with.
