Make Looker Continous Integration work better for you

Key Takeaways

Come on, come on. I only changed one dimension. Why is Looker continuous integration retesting the entire model? I need to get this change pushed. Hey, hey, hey. Don't blame me. By default, I'm a perfectionist and want to test your whole model, but we can definitely change the configuration to speed this up. Hey Looker CI, [music] we already have a team plan to work on older errors for the whole model. This is just for a simple [music] update. Then you need incremental validation. It's a specialized method for isolating errors that are unique to your development branch. It ignores [music] existing errors in production, so you can focus strictly on the ones you're responsible for. It makes validation much faster, especially in large projects [music] with hundreds of explores. That's exactly what I need. Is that the same as targeting an explore? Not quite. Incremental validation is about error isolation between branches. Targeting explores is about scope. To target specific [music] explores, specify the model name, then a forward slash, and then the name of the explore. If you want to exclude an explore, maybe because it's large or someone else is working on testing it, use the same formatting in the explores to exclude configuration option. [music] Okay, incremental validation to isolate my specific changes, and targeting the explore I'm working [music] on to fix older errors because I know I'm responsible for some of those. Now you're focused. No more distractions, just pushing clean code. Finished. And I know exactly which errors are mine.