Let’s Talk About Continuous Integration

Real-world examples of automancery at companies like yours

Start up any automation on a schedule, and create a clean, easy-to-read report on a dashboard that everybody can understand. How simple is that?

50% Reduction of Time in Test Runs

Challenge: A team using Jenkins as their Continuous Integration server found that their test runs were taking 90 minutes.

Solution: Jenkins had access to about 8 “nodes” (individual computers that ran a series of “jobs”). The problem ended up being that each node had a dedicated set of jobs to run, so when a node was finished with its jobs, it just sat there. The solution was to put all the jobs into a big list, and dole out the next available job to the first available node. If no nodes were available, wait a bit, and try again.

Result: The resulting performance increase was dramatic–the test run took 45 minutes instead of 90, meaning that the team could run the battery of tests more frequently, find faults sooner and drive them to resolution, in record time. They were especially excited to be able to start the tests, take a lunch break, and have the results actually be ready when they got back.

 

Scheduling Automation

Challenge: Once an automated script is fully tested and robust, do you really want to sit there running them at a certain time every day? Survey says: nope!

Solution: We routinely use Jenkins to pull down the most recent version of automation, and run it at strategic times during the day. LinkedIn scraping done between the hours of 8a – 4p every business day, email campaigns sent at reasonable hours, and maintenance done during off hours.

Result: Once automation is loaded into the server, it’s set-it-and-forget-it. Much lighter mental lift knowing that things are taken care of on your behalf.