What triggers me most: saying DevOps is a role or a team.

@rEaL-jAsE dammit you :D

I was thinking the same last day... I ofter hear about "devops teams" and I don't get exactly what they should do all day... Maybe someone could enlight me

As one who deals with devops crap all day long. Devops is not easy, when you have to deploy and maintain a large set of client system, each with its own sepcific config.
my case is pretty uniqe:
1. old legacy client set based in an old node version, no ui.
2. old version based on node+docker+docker compose. has two ui varients two clients uses this, each with whitelabel on ui. the system has two feature set.
3. new go+k8s system that is overly complicated and under development.

the job is essetially changing the dev team work to allow easier deployments - setting up configs+deployment sets+optimizing code+version migration
while I support the older clients with legacy updates of old code and automatin, while trying to stay on top of things.

devops can be very interesting - but it not when you maintain 1 cloud based server + frontend ui or a mobile app. did that, there was no need for devops.

We have a dedicated DevOps team of 25+ people. They manage configuration of delivery pipelines, cloud hosting, working with cloud service centre, scaling, uptime requests, enterprise design, working with development teams to help configure their projects etc.

Would be too much for the development teams to do alone.

@LynxMagnus you mean a dedicated Ops team? Or are those 25 people the only ones using DevOps methodologies in their daily work, while the rest of the org does something else?

@ojrask that's how it is in my company. The application developers are strictly devs, with maybe a tiny little bit of Ops work but still mostly devs. Then my team (which has the very-buzzwordy, I-saw-it-in-an-article-once title of "SRE Team" (unrepentantly stolen from Google), works with the developers to set up and deploy their infrastructure (via code/configuration, which is where the dev part of devops comes in), set up deployment pipelines and automation, manage basically the entirety of our cloud infrastructure in AWS and Azure, handle major infrastructure-level incidents (there's the Ops work) and basically do the job of two teams with a team that is too small for one of those roles, let alone both.

From the way I wrote that, it sounds terrible, but I actually enjoy it most of the time when we aren't in the "Oh shit everything is falling apart" mode.

"Cloud Infrastructure and Automation Engineer" would be a more accurate title but "DevOps Engineer" is shorter.

@ojrask dedicated ops but prescribe patterns to the teams.