Mass is a tool for manipulating lots of servers concurrently. Open up an SSH terminal to each server that maches your request, do the same with cluster SSH, upload a file to every server, download a directory into a separate folder for every server, make modifications to them then upload each copy to the right server. Generate a catalog of your infrastructure using the AWS API, and quickly query that using arbitrary searches.
Mass is highly extensible. There’s so much more than I could possibly describe in a couple of paragraphs. I’ve typically been pretty poor at convincing people to try it, yet those that do are very quick to get new people on it “A couple of hours ago I had never used it. Now I can’t imagine my job with out it.”
Get it at github.com/ksandom/mass
And see the documentation at github.com/ksandom/mass/tree/master/docs
Note that in the video I said that I’ve written this tool at every company I’ve worked at for the last 7 years. In a couple of years that will be entirely true, since it was 2007 when I first wrote a tool that was recognisably this concept. However I’ve also wrote key steps that lead to this since 2005 and before. But really 2007 is the key moment when people started to take notice. The point is mass has been through several iterations learning from each that has passed.
PS: There’s a story behind the [debug0] message in the picture above. You’ll understand if you install it ;)