Coding Sheet
Every problem comes from a real DevOps, SRE, Platform or Cloud interview report, so you learn only what shows up. Start from zero, get fluent in Bash and Linux, then move on to DSA patterns.
- Problems
- 251Problems
- Tracks
- 3Tracks
- Sections
- 46Sections
- Companies named
- 61Companies named
How it works
1. Pick your role
Choose Cloud, DevOps, Platform or SRE and the sheet shows only the problems that role is asked.
2. Work track by track
Go in order from programming basics and Bash to DSA patterns, from zero coding to interview-ready.
3. Revise before the interview
Use the recaps and company tags to revise the must-do problems without opening every page.
Why this sheet is different
Built from real interview reports
Every problem comes from a real DevOps, SRE, Platform or Cloud interview report, not from a software engineer list.
Ordered from zero to ready
It starts with Python from zero and Bash, then moves to DSA patterns, so nobody is stuck at step one.
Role-aware
A role filter shows only what your target role needs, so you skip problems you will never see.
Frequently asked questions
Usually yes, but a much lighter version than software engineers prepare for. Most interviews that include a coding round ask easy to medium problems built on arrays, strings, hashmaps and sorting, the same skills you use to count, group and search through logs and configs. Many companies also run a practical round instead of, or next to, a LeetCode-style one: parse an nginx log, process a CSV, or write a small automation script. Hard dynamic programming is rare outside Google SRE and Meta Production Engineer loops, and service companies often test very little DSA at all, focusing on tools and scenarios instead.
Yes, noticeably. SRE interviews have the widest range: at large companies like Google and Meta you can expect medium problems on graphs, heaps, intervals and streams, plus a practical round such as aggregating a log file or calling an API. Platform roles lean towards building small systems from scratch, like an LRU cache, a rate limiter or a worker pool, and Go shows up more often. Cloud roles usually ask easier data-structure questions and put more weight on scripting, cloud SDK automation and networking. DevOps sits in between, with heavy use of Bash. That is why every problem here is tagged by role.
Yes, that is exactly who the first track is for. Most DSA sheets assume you can already program, which leaves people from support, sysadmin or non-CS backgrounds stuck at the start. The Programming from zero track teaches Python basics first, using infra examples like server lists, log lines and config values, so you learn the language and the domain together. From there, Bash and Linux covers the shell, and only then do you move on to DSA patterns. Following the tracks in order takes you from no coding experience to interview-ready without gaps.
Those sheets are excellent for software engineering interviews, and they are built for them: far more problems than any infra role is asked, including hard dynamic programming and advanced graph topics that infra interviews rarely touch. This sheet keeps only the DSA patterns DevOps, SRE, Platform and Cloud roles are actually asked, then adds what general sheets leave out entirely: Python from zero and Bash one-liners. A role filter shows only what your target role needs, so you don't spend months on problems you will never see.
The full sheet has 251 problems, but no single role needs all of them. With the role filter on, Cloud comes to about 182, DevOps about 250, Platform about 150 and SRE about 251, and one problem can count towards several roles. At a steady pace of two or three problems a day, most people can work through their role's list in two to four months, faster if they already code. You can also skip tracks you're already confident in, for example Programming from zero if you already write Python.
Use Python for most problems. It is the default language in DevOps, SRE and Cloud interviews, its standard library covers what practical rounds need (files, regular expressions, JSON, CSV and HTTP), and interviewers rarely object to it. Use Bash for shell tasks like grep, awk and find pipelines, which several companies test directly. Go is increasingly common on Platform and infrastructure teams, so if you're aiming for a Platform role it's worth solving some of the design-heavy problems in Go as well. If you already know another language well, the DSA problems still work in it.
Yes, a bit more of it. Google's SRE software-engineering track and Meta's Production Engineer loop hold coding rounds close to a software engineer's, typically medium problems written without running the code. The Advanced track covers the extra topics those loops add: Dijkstra, union-find, a median of a data stream, harder sliding-window problems and core dynamic programming. Work through the DSA patterns and Advanced tracks with the SRE or Platform filter on, and practise writing clean code without an editor's help. For most other companies, the core tracks are enough.
Yes. Like everything else on DevOps Network, the Coding Sheet is completely free, with no paywall and no sign-up needed to read it. A free account lets you tick off the problems you've solved and see your progress per pattern and for your role.
Get notified
All tracks are live. Leave your email to hear when new problems, explanations and video solutions are added.
Join the Network
Get advanced DevOps guides in your inbox.
Sources
Based on candidate interview reports (Blind, Glassdoor, GeeksforGeeks, LeetCode Discuss) and these public interview guides:
- DevOps/SRE coding interview levels (Reliability Whisperer)
- Google SRE interview questions (Coditioning)
- Meta Production Engineer interview guide (Coditioning)
- AWS Cloud Support Engineer interview guide (Prepfully)
- Hiring SREs at LinkedIn (LinkedIn Engineering)