@chrismessina

Thank you, an excellent writeup.

There are definitely a bunch of points here for anybody trying to keep an internet service up. The ones that jumped out at me were:

The contrast between responsiveness expectations formed in corporate life versus just how fast a hobbyist instance admin can respond.

How admin contacts on a service and admin access to the underlying mechanisms can't be dependent on that service, like how companies run shiny.example but the employees are all on shinycorp.example and the clusters are like cluster1.shineinfra.example.

The service-breaking potential of abuse reports to upstreams when your thing is hosting (which many internet presences are).

The lack of patience and rote responses from providers whose expectations have been formed by decades of abusers and bad faith interactions before they've even heard of you.

The limited number of notification options when "unexpected burst of volume from new source" and "unexpected pattern of traffic from an existing source" are lighthouse signals of spam messaging.

All the long tail stuff (monitoring!) is really difficult as a sole proprietor.

There's definitely more in there and many people should read it. Heck of a confluence of things you just had, definitely.