TIL about Concourse fly targets permalink

At work we use Concourse CI for continuous integration and delivery. It encourages interacting with it via the fly commandline interface (CLI). It also requires everyone to log in to it, and be a part of a team, to do a form of role based access control (RBAC). Today, I lost about 30 minutes trying to figure out why my new pipelines weren’t being deployed to the correct team in Concourse. It was because I had set up my fly target incorrectly.

A target is essentially a separate fly session. When you log in to the fly CLI, you specify

• A target name underwhich to store the session creds
• The Concourse CI server URL at which to log in
• An optional team name to log in as

If you don’t specify the optional team name, it will use the default.
In my case, our default team is main.

Then, whenever you try to deploy your pipeline like so:

fly -t $target_name set-pipeline --pipeline my-awesome-pipeline --config my-awesome-pipeline.yml

The pipeline is deployed to the team space in Concourse for the team specified by your target.

You can view all your configured targets using the fly targets command:

λ › fly targets
name       url                                      team       expiry
rgmt       https://concourse.test.robgmills.com     test       Sat, 25 Apr 2020 00:25:24 UTC
rgm        https://concourse.live.robmills.com.com  admin      Sat, 25 Apr 2020 00:25:24 UTC

You can delete and edit targets:

λ › fly -t rgmt delete-target
deleted target: rgmt
λ › fly -t rgm edit-target --team-name live --target-name rgm-live --concourse-url https://concourse.live.robgmills.com
Updated target: rgm

I needed to learn this so I could deploy a new pipeline at work. In the process, I killed two birds with one stone: added a CD pipeline that terraformed our Auth0<>SumoLogic integration and user acceptance tested one of the new CI/CD features my team built.