Best Practices

config.yaml

These custom overrides are good for a dedicated Continuous Integration server:

---
pidfile_path: /var/run/robot_sweatshop
logfile_path: /var/log/robot_sweatshop
job_path: /etc/robot_sweatshop/jobs
workspace_path: /etc/robot_sweatshop/workspaces
database_path: /var/db/robot_sweatshop
scripts_path: /etc/robot_sweatshop/scripts

Async

A worker will go down the list of commands and execute them in a blocking manner. However, you might want to kick something off in a non-blocking way. Since there is no limit on the jobs you can run, I suggest branching jobs.

For example, job A does work and generates the API from the compiled code before it runs tests. Generating the API and running the tests both take a while and they have no dependency on each other. Create job B which runs the API generation and replace the command in job A with a curl POST that kicks off the new job. Then both commands will be running at the same time on different workers. You can chain as many jobs together as is needed in this manner.