Who Needs Site Reliability Engineers (SREs)?

From its humble origins as a role inside Google, Site Reliability Engineering has become a type of position or team that a wide variety of companies now embrace.

But that doesn’t mean that every company under the sun needs SREs. In this article, we unpack how to determine whether SREs should be a part of a given organization. In so doing, we aim to help both employers who are trying to figure out whether they should invest in SREs, and SREs wondering which types of companies are looking for the skills they stand to offer.

Why are SREs important?

Understanding who needs SREs starts with understanding why SREs are important to companies.

If you are an SRE, or you work with SREs, you already know the answer to this question. The main benefit of hiring an SRE (or SREs) is that businesses gain engineers who specialize first and foremost in ensuring that software works as it’s supposed to. In turn, SREs help to keep customers happy, while also mitigating the risk of disruptions to the systems businesses need to manage internal operations.

SREs also help in some ways to bridge the gaps between other types of roles. It’s important in this respect not to conflate SRE with DevOps, which more explicitly integrates IT with development. But by the nature of their work, SREs also help developers, IT engineers, QA engineers and even non-technical stakeholders (like customer relations teams) to collaborate.

Which companies should hire SREs?

Although the SRE role existed only inside Google for most of its history, that has changed over the past five years or so. Today, most of the big-name tech companies, such as LinkedIn and Netflix, proudly boast about how they have built SRE teams.

But Google, LinkedIn (which is owned by Microsoft) and Netflix aren’t representative of the typical company. Although SREs might be nice to have for every company, justifying the effort and money required to build an SRE team hinges on whether conditions like the following are true:

• You have ultra-strict reliability needs: Companies like Netflix need to meet very stringent performance standards. After all, even blips of just a few seconds can be a big issue with video streaming. But if most of your user-facing tech is stuff like standard websites, ultra-high reliability and performance may not be as critical. Again, it would be nice to invest in these things if you can, but some companies have more to gain than others by hiring SREs to optimize reliability.
• You make strict reliability guarantees: Similarly, SREs matter more for businesses that need to meet strict reliability standards that are defined in SLAs and SLOs. If reliability issues could trigger contractual penalties, you may want to invest in SREs to help avoid those issues.
• You have large development and IT teams: The larger and more complex your development and IT organizations, the more you stand to gain by hiring SREs who can ensure that reliability remains a priority across these sprawling teams. Smaller organizations with just a handful of developers and IT engineers can often enforce reliability goals without dedicating a role to them.
• You have a complex technology stack: Managing reliability is simple enough if you’re dealing with monolithic apps running on bare-metal servers. With a simpler stack, there are fewer things that can go wrong, and fewer variables to evaluate when trying to optimize for reliability. SREs become more valuable for companies with complex environments, like microservices-based apps.
• You have a continuous delivery pipeline: Along similar lines, SREs are more important for businesses with complex, fast-moving software delivery processes. If you are only updating your apps a few times a year, you probably don’t need SREs as much as a business that pushes out updates daily. In the latter case, SREs help ensure that software remains reliable even as it constantly evolves.

In general, these points could be summarized as follows: The larger your company, and the greater use it makes of modern, fast-moving technology and technological processes, the more you stand to gain from hiring SREs.

Conclusion

SREs are never a bad thing to have on your team. Devoting resources to SRE will never hurt.

However, the reality is that some businesses need SREs more than others, and some will yield a higher return than others on the effort and money they invest in building an SRE team. Companies and SREs alike should carefully evaluate these factors before committing to an SRE role.

{{subscribe-form}}