The Only Constant Is Change: Lessons from a 25 Year SRE Career

Twenty five years ago, when I started as a sysadmin and attended my first USENIX conference (LISA 1998), "SRE" didn't even exist as a field, and the compute infrastructures we supported were far smaller and more localized. Now, SRE is an industry standard job across the world, and SRE's manage massive and extraordinarily capable cloud environments distributed around the globe.

But the pace of technological change is not slowing down: with rapid development in new technologies such as Machine Learning and Large Language Models (e.g. ChatGPT), we can ask ourselves: will SRE still exist 25 years from now, and if so, what will those jobs be like, and what skills will they require? We cannot know the future, but this talk will concentrate on the things that we can know: job and career skills that have served the author well over the years, and cannot be easily -- if ever -- replicated by large volumes of compute power or any foreseeable artificial intelligence.

Using examples from the author's career through the industry, starting from maintaining a handful of systems in small organizations, all the way to the highest engineering ranks of one of the world's largest tech companies, we will discuss what has worked well, what career options are available, and how to keep your job skills relevant in a world that is constantly evolving.