[Andys]

From my perspective, its a continuing boom over decades.

New abstractions come along (eg. AWS), bringing additional complexity, then better tooling gets developed to deal with it, and this prepares the way for new abstractions on top (eg. Kubernetes).

At the same time, legacy systems remain around a long time and need experienced people to maintain those.

While operators are glad they don't have to deal with bare metal any more, there's a lot of additional complexity you need to understand to manage AWS effectively, that didn't exist for most companies back in bare metal hosting days.

[Gigachad]

The complexity doesn’t seem to have really moved that much, just the output has absolutely exploded to the point where the average dev ops person is expected to do things that only tech megacorps would have pulled off 15 years ago.

It’s no longer acceptable for anyone to turn their service off for maintenance for example. They all have to roll updates out frequently and without downtime. They all have to scale to huge numbers of users and data. They all have to have almost perfect uptimes.

[tester756]

There's online game, which runs since 97 and has profits around 20 milions of EUR

and they turn off their game everyday for 10mins :P

[rightbyte]

... what game? Runescape?

[tester756]

Tibia

[sjaak]

tibia?

[tester756]

yup