Hi. I'm Suresh - DevOps/automation engineer, 10 years in, currently deep in AWS, Azure, Kubernetes, Terraform, and Ansible for a living. I joined DEV two weeks ago. I've published two articles. I have no idea what I'm doing with this platform yet, and I'm having a genuinely great time finding out.
This isn't a "here's what I learned" listicle. It's more like - pull up a chair, let me tell you what the last two weeks looked like from the inside, because some of it was funnier than I expected infrastructure writing to be.
Why start now, after 10 years of not writing
Short version: I got tired of solving the same problem quietly, for the fifth company, and having literally nothing to show for it except my own memory. Thirteen years of pattern-matching - "oh, this is the local-state-on-a-laptop problem again," "oh, this is the giant-shared-state-file problem again" - and none of it existed anywhere I could point to. It just lived in my head, useful to exactly one person.
So: write it down. Worst case, nobody reads it and I've still got a clean writeup of things I already knew. Best case, someone googling the exact 2am problem I've already solved five times finds it and skips the pain.
What actually happened in two weeks
I built a self-healing Prometheus + Alertmanager pipeline that restarts nginx on its own, wrote it up, and - this is the part nobody warns you about - spent longer getting the cover image and Dev.to formatting right than writing the actual article. Nobody tells you that going in.
Then I wrote about the five Terraform state mistakes I've watched happen (and caused, once, early on - we don't need to relive that one in detail). Somewhere in the middle of publishing that one, I also deployed a small free tool, and this is the actual highlight of my two weeks:
The deploy got stuck. Queued. For twenty minutes. Then I retried, and it queued again. Turns out it was a known platform-side issue completely outside my control - nothing wrong with my file, nothing wrong with my setup, just a stuck queue with no ETA. Thirteen years of infrastructure experience, and my solution was the same one I'd tell anyone else: stop waiting, fail over. Dragged the exact same file onto a different host. Live in about ninety seconds.
The irony was not lost on me that I'd just published an article about self-healing infrastructure the same week my own deploy needed a very manual, very human intervention to heal itself. Infrastructure has a sense of humor. It is not always kind.
The actual surprises, ranked
- Writing the article is maybe 40% of the work. Cover image, tags, formatting, screenshots, re-reading it four times to make sure it actually sounds like me and not like a template - that's the other 60%. I did not budget for this.
- People are nicer than I expected. I came in braced for the stereotypical "well actually" comment section. Hasn't happened yet. Might just be early days, but so far it's been genuinely constructive.
- The "did I really live this or does it just sound made up" test is real. I made myself a rule early on: if I can't answer a specific follow-up question about something I wrote, I don't publish it as my own story. Turns out that's a surprisingly high bar when you're staring at a blank editor at 11pm.
- I have way more stories than I thought. Thirteen years apparently generates a lot of "oh, that reminds me of the time..." - I just never had a reason to write any of them down before.
Where this is going
More of the same, honestly. Real problems I've actually hit, real fixes, occasionally a tool if I get annoyed enough at doing something manually more than twice. No hot takes on whatever's trending this week unless I've genuinely got something to add.
AMA, kind of
Two weeks isn't enough runway for a proper "ask me anything," but I'll happily answer anything about:
- DevOps/platform engineering as a 10-year career (AWS, Azure, Kubernetes, the lot)
- What it's actually like starting a technical writing habit from zero, in real time, badly, alongside a full-time job
- The self-healing pipeline or Terraform state posts, if you've read either
- Whether infrastructure really does have a sense of humor, or if I'm just tired
Drop a comment - genuinely, I'll answer everything. Two weeks in, this is still the part I didn't expect to enjoy as much as I do.
Top comments (3)
I can relate to this. Writing the article often turns out to be the easy part—editing, formatting, screenshots, and making sure it actually sounds like you usually take much longer than expected. 😄 The good thing is that every article becomes a reusable piece of knowledge instead of another lesson that lives only in your head.
Thanks @merbayerp
Exactly this - "reusable piece of knowledge instead of another lesson that lives only in your head" is a better way of putting it than I managed in the whole article, honestly. That's really the whole reason I started😃
Glad it resonated! Looking forward to reading more of those real-world stories.