The Shite Reliability Engineering Manifesto

• We all want to get code to production quicker, so why not take Continuous Delivery to its conclusion and vim on the server?
• Canary Deployments are great. Half measures are not. Go 100% or go home.
• Every advancement since the 1970’s has been a waste of time. There are no computer science problems in existence that cannot be solved with a combination of sed & awk
• Always make sure you containerise everything and use advanced scheduling software. This way you can make sure your infrastructure is ready when that static website suddenly becomes a complex financial analysis platform.
• Metrics such as SLIs, SLAs and SLOs are great to have productive discussions. Throwing more instances in is usually better. Follow your primal instincts.
• Alert fatigue? Send those alerts straight to your fitbit and never miss a beat again.
• Start the weekend on Thursday to avoid those no deploy Friday’s