Ever have a code release go horribly wrong? Have a routine system upgrade turn into 12 hours of downtime? Had to field angry phone calls from engineers, customers and your boss?
Sometimes things go horribly wrong. This talk will teach you how to plan for the worst, minimize risk and recover gracefully from failure.
In an ideal world, code would have complete test coverage, releases would be seamless and databases wouldn’t require downtimes to upgrade. The fact is, most companies still have to manage systems upgrades and code deployment that includes downtime. This talk is for you.
Elements to successful downtimes include: proper testing, automated checks for success, completely failback plans, and a timeline. We’ll go over useful tools for keeping track of what is going on before, during and after a downtime, and talk through good and bad communication for employees and customers.
We’ll also talk about how non-technical people can evaluate technical change. And how having the right stakeholders involved from the beginning of the planning process reduces risk.
Stories to illustrate success and failure will come from Selena Deckelmann’s 15 years in education, public and private sectors, as a consultant, DBA and IT manager.