In my experience its true. When S3 went down my companies systems started malfunctioning, as well as many other vendors/system all over the internet including things like slack. Our customers were experiencing pain from multiple vendor failures. When our customers cant order lunch, run trello, or sent chat messages on slack they blame "the internet".

Note:I take offence to being called an AWS devotee, I have been in this space professionally for going on 13 years with nearly all of it in the startup space. The early years had to rely on collocation. To see startups struggle from things like hardware failure and delaying sales because they need more hardware is something I do not fondly remember.

Apparently “the internet” to you is just hip startups if your name call outs are things like slack, trello, and lunch ordering. The startup echo chamber and the AWS echo chamber have a lot of overlap due to the elasticity you highlighted. That still doesn’t mean it’s representative of internet services.

When an AWS outage stops wire transfers, airline reservations, robot surgeons, etc, then it really will be “everyone”. Until then, try to give yourself some perspective so you can recognize when a product is mature enough that it’s time to start moving away from AWS.

> I have been in this space professionally for going on 13 years with nearly all of it in the startup space

This is why the siren call of instant infrastructure is so alluring to you. While your depth of knowledge for startup infra requirements is there, it does not transfer to large enterprises/campuses/etc where demand is very predictable and involving another company (Amazon) in your operations is nothing more than a liability.

Before you bemoan on-prem competing with Amazon’s world-class engineering org, consider that their priorities are spread across thousands of products all interacting with millions of customers. An outage that could destroy your business means nothing to them other than some reduced KPIs that month. While an on-prem solution might be worse in throughput, global latency, etc, the fate is in control of the org that can make sacrifices Amazon can’t (e.g. planned nightly outages).

* This is why the siren call of instant infrastructure is so alluring to you. While your depth of knowledge for startup infra requirements is there, it does not transfer to large enterprises/campuses/etc where demand is very predictable and involving another company (Amazon) in your operations is nothing more than a liability.*

Every large company has dozens of vendors that they “involve their operations”.

The best part about companies that use the cloud is they don't have to deal with joyful sysadmin gatekeepers like you. That is like 75% of the reason most people use the cloud. To avoid dealing with gatekeepers such as you who "know better". The ones that slow every god damn project in the company down...

> This is why the siren call of instant infrastructure is so alluring to you. While your depth of knowledge for startup infra requirements is there, it does not transfer to large enterprises/campuses/etc where demand is very predictable and involving another company (Amazon) in your operations is nothing more than a liability.

My whole original post was from the startup point of view and I made that quite clear. I am more then happy to admit the enterprise space (from an infrastructure POV) is not my expertise. Your taking my points out of context. Even in my own example I added an astrisk of a startup who started moved to hybrid on-prem after six years.