I am reading the book “That will never work” by the founder of Netflix. From a technical point of view, something that really intrigues me is their launch process. The engineering team proposed an increase in server bandwidth to accommodate a higher request volume; however, the business unit expressed uncertainty regarding the projected demand. This result in a lot of guess work and a huge stress on the engineering team during the launch day. On the positive end, they did have stage and prod environment, which is a best practice approach. Scalable system architectures are now critical, yet comprehensive load and performance testing remains insufficiently prioritized. Similar things happened to Healthcare gov on it’s launch day. The key takeaway is the importance of proactive planning and scalable system architecture to accommodate fluctuating demand.
The daily journey by Aleks Vladimirov