This all seems very confusing.<p>The Cloud was too expensive, so they saved money by just buying servers up front without talking about the hosting costs.<p>They didn't need to hire more people because all their folks are doing sys admin stuff. Oh except for the fact that the actual reason they didn't need to hire was because they have outsourced all the hardware management.<p>Sure, it really might be cheaper, especially with egress considering their products, but there is a lot of caveats that are getting glossed over here.
I have the vague feeling that way to many manager with poor technical skill make decisions like "We're going to the cloud. All-in!" or "Use a cloud-native™ approach!" without understanding all implications or to tackle problems (like fossilized business processes) that have nothing to do with the server infrastructure as such.
> No, because we didn’t change the team composition after our cloud exit.<p>This has been my experience with my major enterprise company’s cloud transition as well. The ops payroll has only grown since the transition. And our company had 4 major platforms before transitioning to the cloud and now only has 3.<p>I won’t deny the cloud doesn’t have any benefits. Distributing services and DBs geographically is much easier than with Colo for example. But really yhwts where the cloud benefits really lie. Since everyone is spending all their money building services for the cloud, that’s where innovation in tech lies right now (although whether we need more innovation is indeed arguable, and whether one benefits from being at the vanguard of innovation is also an open question).
Related last week:<p><i>Leaving the Cloud: Cloud Computing Isn't for Everyone</i><p><a href="https://news.ycombinator.com/item?id=38644550">https://news.ycombinator.com/item?id=38644550</a>