Tiny struggles

Always hacking something 🧑‍🔬.

I left my Site Reliability job

I left my Site Reliability Engineering job after almost seven years. I didn’t leave Google though, just changed to another role within Software Engineering in the Cloud Infrastructure.

I was somewhat unhappy with my cushy role as a Site Reliability Engineering (SRE) Manager at Google. After quite a lot of deliberation, exploring opportunities and some setbacks (hello, industry layoffs and hiring freezes), I left my SRE job. My last day was October 14th 2022.

I was in SRE for almost 7 years even though I never actively planned SRE to be my career.

How I ended up as an SRE

Let me say this, even though I have worked in Backend Development positions pre-Google, I had interests that would make me a good SRE candidate.

In 2014/2015 I learned about DevOps. I set up all automation and deployment in the startup I worked at the time and I got fascinated by containers and all things cloud infrastructure.

I wrote multiple blog posts about Docker, gave workshops, organized meetups and even had a book deal (seriously! They give book deals to kids like me at the time!). I also was surrounded by various hardware hackers and serious sysadmins in the Warsaw Hackerspace. Hackerspace was my second home, and I would spend 20+ hours per week there hanging out with other hacker friends, building various software and hardware projects.

Even though I was very successful as a Dev in Poland, when Google offered me an internship abroad, I jumped at it. Being 22, it was a great opportunity to get out of my comfort zone. I guess because of my DevOps/Infra interests I got slotted into a SRE internship.

SRE internships are a bit weird, you basically end up being a Dev within a SRE team. It went pretty well and I got excited about the opportunities, the tech was so cool, people were really smart and inspiring. I wanted to try working for Google long term.

And I got it. Even though Dublin was not my first preference I accepted the role and ended up in Dublin as a SRE. I mentioned to the recruiter that I would prefer the SWE position, but somehow “it wasn’t possible”. Well, I was young and didn’t push for what I wanted hard enough. I knew that coming from a SRE internship it was a default path and was just happy about the opportunity in the end.

Inertia and SRE only site

Later on I learned that Google Dublin was an “SRE only site”. All the eng teams were in SRE.

I uprooted my life to move to Dublin, which was disruptive on its own. But later I also added a lot of inertia by:

  • accumulating significant investments
  • buying an apartment (1 & 2)
  • getting married here (to an American - creating a international couple leaving abroad)

All of these would make a move quite a hassle.

Ireland is not the best country in the universe, but it’s pretty damn decent. It has a lot going for it that is aligned with what I value and enjoy.

Well, why not leave Google and get a better suiting job in a different company?

Here are some reasons adding to the inertia:

  • To interview you need to have enough mental space and practice interviewing.
  • Google pays very well compared to other companies in Dublin, offers great work life balance and perks (free food, subsidized massages, etc)
  • After a while my outside of Google options would be SRE or management roles because this is what I was doing currently and had experience in - which if I was to do, I would rather do at Google

It was much easier to just continue and make the most out of it, instead of making a big jump into unknown.

Making the most of it

Even though that I would have preferred to be a SWE instead of SRE, the job gave me an opportunity to work on very interesting engineering problems and codebases with some amazing engineers.

I learned how to do “good enough job” operationally and focused on making impact in a way it was expected from an SRE.

In infrastructure products, most of the difficult, design and code heavy engineering falls onto the dev teams, not the SRE teams. In some ways, SREs shield the dev teams from disruptions and cognitive load to make progress on cranking out features. However, SRE work can still be pretty interesting and I was set on making the most out of it.

I pretty quickly gravitated to leadership roles. Being given trust and autonomy makes any job easier. I quickly realized that many of the realities that we take for granted are just a current setup, which is not the only possible one. The rules and dynamics can be changed. Vision and direction will often stem up from the personality and opinions of the given leader or set of leaders.

I was on a roll, I was productive, I got promoted pretty quickly (L3 -> L4 in 9 months, L4 -> L5 (senior) in 1.5 year, L5 -> L6 (staff) in 3 years).

When things become somewhat unbearable, e.g. I got bored of the type of the work I was doing and types of problems available, I would switch things up. Being promoted helped with this as I was given bigger and bigger opportunities.

I made 3 job changes before I left SRE, Individual Contributor (IC) -> Tech Lead (TL) in my first team, first team to the second team and then in my second team from IC to Manager.

Each time the change energized me and triggered a lot of growth. Changing teams helped a lot. Leaning into leadership empowered me and taught me a lot even though it usually was more stressful than IC roles.

Skill growth

Growing skills is important for building confidence and increasing your value to the team, but also it can be satisfying by itself.

I derive a deep satisfaction from learning. And ideally, I want to keep my personal skill interests and job interests aligned as much as possible.

And I did for a while, learning more about distributed systems, databases, reliability was pretty interesting. (At least for a while, but it’s applicability to my day to day job was limited).

When I decided to switch to management, it unlocked a lot of new learning. It got me out of my comfort zone and forced me to become more mature.

I experienced new organizational dynamics and people dynamics. I had attrition, I hired, I trained, I managed performance, I motivated, I coached.

It pushed me to learn more about myself and other people. I researched books, I read and read, and read. I learned a lot of soft skills and how to be more resilient. But I think I reached a point of being pretty good at my job in that specific environment.

At the same time, my technical learning at the job slowed a lot and the opportunities for learning were not that exciting to me.

Overall, I’ve never been passionate about being very good at troubleshooting, or managing incidents or crises. Probably I’ve been more interested in that than a typical Dev, but comparing to some of the best people I knew, the difference of interest was clear.

In the meantime, I’ve been doing quite a lot of indie hacking and building products from scratch. This also triggered desire to learn, but in many ways it wasn’t aligned with what would help me get ahead at work.

It caused me feelings not having enough time to learn all the things I wanted to learn as my interests weren’t aligned.

Stress, overload, organizational tensions and multi-year projects

I didn’t change jobs because of burnout, but I had moments of burning out before I made the decision to switch.

There was stress, overload and issues. This winter I had some work drama with a poor performer, there were problems with hiring in Dublin and some pushy managers of other teams. On top of it I was training very hard and working on side hustles. It didn’t feel to me as if I was working on my job that hard, but the stress was amplified by me putting myself under other demands and also a fundamental feeling of misalignment.

Side projects made it very clear to me that I need creative outlets, that I want to build my own company, create things from scratch. Meanwhile I had to go to another meeting, write performance reviews and do another oncall shift.

Dev and SRE are supposed to be partners, but often have a tense relationship. In a typical product area, Devs are leading the org shipping new features, making existing infra better and the SRE are exposed to risk and toil coming from these changes. SREs have their own projects, of course, but need to juggle lots of demands and dooming risks.

If you ever heard about circles of concern and influence, a typical SRE TL or manager would have a very wide circle of concern, wide circle of influence and tiny circle of control (due to limited staffing and other commitments). Of course influencing works, but it’s not free and has significant time and cognitive cost. I didn’t enjoy this dynamic too much.

Additionally, I felt stuck with a very long term project. It’s been multiple years in the making, I impacted it a lot and was deep into the weeds. It would keep evolving for many years to come, but wouldn’t be fully done for many years. It was still exciting in many aspects, but the rate of learning has decreased significantly. I felt like it would be pretty hard for me to get untangled from it.

Overall, I realized that this job was making me unhappy and unsatisfied.

I had a lot of decent days, some amazing days, and some “oh crap, I don’t want to do this anymore” days. It’s possible that my unhappy days would just be unhappy and I would find a reason to justify it. Human minds and emotions are not to be fully trusted! It’s also possible that things like my side hustle being a crappy business contributed to some negative emotions.

Finding alignment & imagining the next role

I decided that I will change roles back in May and planned to end my SRE role at the end of September. The team wasn’t in the best place, there were projects I wanted to leave in a good place and I wanted to explore the available options.

I wasn’t exactly sure what I wanted. I knew what I didn’t want - management and SRE.

Pandemic opened more opportunities and people were more open to hiring remotely which was quite exciting.

I dated a bunch of different opportunities. Would project management be too big of a step? Maybe Chrome, it would be so different from SRE. DevRel could give me some useful experience building products intended for developers.

I even bought a book on finding the job to love (which I still haven’t finished!).

I wanted a job that would be better aligned with my long term goals (entrepreneurship and building stuff) or generally a low demand job that would let me focus on those other goals in my free time.

My self imposed deadline was getting closer and I still haven’t found the right job when the internal hiring freeze hit.

Let’s say that it messed up some of my plans. But in the end I ended up with a role that was both a great match for my existing experience and for what I did next. So I guess the long search time paid off?

Leaving good things behind

Even though, the job wasn’t aligned with my long term goals and was making me unhappy, it was hard to change.

As a manager I had my team. I developed pretty deep relationships with people who I managed and also had a great relationship with my own manager and with many other partners across the organization.

The problems with overload also drastically improved over last couple months and in general problems were being addressed.

I was also good at my job and could make solid impact with relatively little effort gliding on my existing knowledge and influence.

It wasn’t terrible. Was I falling a victim of “grass is always greener”? Could the new role be significantly worse?

What next

Directionally, I still want to build a company at some point. I want to grow as a leader. But I also want to grow as an engineer and be a very skilled individual who understands math and science. It’s hard to get all those things at once. For now I will focus on my love for engineering.

Should I put off building a company for another couple years if I wanted to do it since I wanted to be a teenager? Through indie-hacking learned that building a product that people want is much more than just building software.

Could I just do it now? Yes. I think that eventually I would succeed.

Working on indie products made me appreciate employment more

Failing to find significant traction with watchlimits (my indie product) made me appreciate how wonderful paid employment can be.

You get paid a lot, you are expected to be learning, the work life balance is good, you have support network and you work in a space where there is already a lot of traction and your work is almost guaranteed to have impact.

So, since I found a role that is better aligned with what I want long term and has the best what paid employment can offer I will try to make most of it.

I will be working for learning and fun from now on

I am financially in a very good place. I don’t have enough to never have to work again in my life, but I’m not that far off. And I am confident that I will earn money as I enjoy working for the most part.

From now on, my guiding principle for work is that I will try to make my job so good that I would want to work there for free. Main reasons for working being: learning, fun and satisfaction of making positive impact.

I ended up picking up the most intense role available to me, challenging on many levels, but I have confidence that I can pull it off without sacrificing my sanity or health. I’m in my prime years and ready for a challenge.

The future is pretty exciting.

Check out my latest project Watchlimits
Watch time insights and limits that work for you!
This is my mathjax support partial