Requirements: an existing Google Cloud Platform account a Git or Github repository containing your project your project can have a Dockerfile (but not mandatory) With Cloud Run, you can easily deploy a container image and let it scale up and down as needed, in a serverless fashion: No need to focus on infrastructure (provisioning servers, clusters, upgrading OS, etc.) Your application can scale transparently from 0 to 1, and from 1 to n (no need for a pager when your app is featured on Hackernews) You pay as you go, proportionally to the usage If your project is hosted on Github, for example, how can you help users get started with your project? Read more...

When I joined Google last June, I discovered a new world: tons of new acronyms or project code names to learn about, but also a particular environment for your source code. At Google, engineers work on a huge monolithic source code repository comprising of: 1 billion files 9 million source files 2 billion lines of code 35 million commits 86 terabytes of content 45 thousands of commits every day. Rachel Potvin, who’s an engineering manager at Google, wrote an article for ACM about how Google handles such a huge repository, as well as the tools and practices around that. Read more...