We (Release-Engineering-Team) had a project last quarter to create a staging cluster, basically a new and improved version of the beta 'deployment-prep' cluster. The project got put on hold because it turned out to be a huge undertaking and we didn't have the resources to get everything finished.

If some other teams want to help pick up this project and see it through to completion, we certainly have made a lot of progress already and that work shouldn't left to languish for long.

In T104324#1498654, @mmodell wrote:

If some other teams want to help pick up this project and see it through to completion, we certainly have made a lot of progress already and that work shouldn't left to languish for long.

See https://phabricator.wikimedia.org/tag/staging/

We did a naive attempt at T104324 with the reading maintained extensions (
http://reading-web-staging.wmflabs.org/wiki/Main_Page) which is obviously
easy to set up. Obviously setting up a staging environment for all the
organization is a huge task :(

I would definitely love to see this tackled as an organizational level, and
having each department owning setting up their respective software and
helping in the process.

We don't have much specific knowledge on ops but if there's something
specific we can help with please ping us.

We're trying to kickstart several efforts to improve infrastructure, for
example T104561 for setting up a common way for teams to specify sets of
test data to import into instances.

ATM we're "hold on" in actually splitting the branches, we'll continue to
talk and consider it moving forward.

Thanks for your comments.