We're about to embark on writing our own downtime/maintenance window scripts to control our Jenkins master...any ideas/suggestions/best practices??

[Gavin Goodrich]

Our IT dept LOVES to take down random parts of our network and infrastructure every other week. This usually means we have builds that break and nodes that are left in a bad state and need manual intervention. Usually IT tells us that this maintenance is coming, so we would like to be able to tell our Jenkins master that we need a downtime/maintenance window during a window of time. For shorter jobs, no problem...just use the api for quiet down...but for longer jobs (we have some that are >9 hours) that are already running this does not solve the problem. We would like to write a script/plugin to do a sweep of the jobs on the server and calculate a "better" quiet down start time.

Before we do, I wanted to ask the community if there is anything like this already out there or if you have written scripts that do this. Anyone??  If not, do you have any suggestions/gotchas/best practices we should be aware of before we start?

Thanks

-gavin