So you just learned Docker and Linux, and while exploring potential job opportunities, you see that SRE is a booming field. It has good demand and is not saturated compared to other software domains like web dev and backend. You see the job description and notice that the tech stack is not that overwhelming. You need to know Linux, Docker, and a few backend frameworks. Cool, you are totally competent to do this job.
You apply and never get any response :(
Compared to other software engineering domains, SRE and Production Engineering are the two fields where the Dunning-Kruger effect hits hard. In web dev or backend, all you need to do is learn a few frameworks, make some projects, maybe look at some open-source production-ready code to see what good industry practices are, and you are good to go. You may still not have any industry experience, but you do know most of the tech stack, and it should not take you more than 5 months to start shipping features.
SRE and Production Engineering are like firefighting. You can read all the books you want, but unless you have enough experience barging into a burning house, you are not going to do well.
This is what makes most entry-level developers not fit for SRE. Having worked 3 years in the SRE and Production Engineering field, I can say most production outages can be solved by running 2–3 basic Linux commands, but it takes years of firefighting experience to figure out what to run and where to run it.
When something fails in production, there are 10 different things that could be responsible. Modern software tech stacks consist of 10+ microservices, a cache, monitoring tools, multiple databases, and load balancers. Where do you even start to look for a problem? This is a classic reference to the Old Engineer with a Hammer story!
Most of the really sharp SREs I've met in my career are veteran developers and sysadmins. They've pretty much worked on all the tech stacks needed in modern production systems.
So how does an entry-level dev get into SRE? The answer is you don’t. You get enough development and DevOps experience to pave your path toward an SRE role. You need to know enough about software engineering and architecture to be able to imagine the entire production system in your head.
For anyone who is still keen on learning how they can work on real-world SRE problems and want to try out their Linux skills, sttrace.com is the right place. It has lots of real-world problems simulating production outages and, in general, the operational tasks an SRE would do. It is also a good place to challenge yourself to learn more about Linux.
Top comments (0)