> That's the whole point. Until something falls over and you haven't deployed any new code and all the vendor's systems are green. Part of the reason you went

> That's the whole point. Until something falls over and you haven't deployed any new code and all the vendor's systems are green. Part of the reason you went with a PaaS in the first place is that you didn't want to manage the infra yourself, you just want to ship application code. This has worked great so far and so you've stopped thinking about the infrastructure at all. Disk and network I/O still exist though, you've just been incentivized to stop thinking about them and so all you know is that the vendor sucks, when in reality it's the application's fault. Somebody's gotta roll up their sleeves and see what's going on in there though, so hopefully your PaaS vendor is as cool as we are :)