Who should I follow to learn about Systems performance engineering?

If you want to get genuinely good at **systems performance engineering**, I’d follow a fairly small group rather than trying to follow hundreds of SRE/performance accounts. ### The people I’d prioritize 1. **Brendan Gregg - start here** - Probably the single best person to follow for systems performance. - Linux internals, CPUs, memory, I/O, networking, profiling, tracing, eBPF, benchmarking, flame graphs, and performance methodology. - His work explicitly covers the full path from applications → OS → kernel → hardware. [Brendan Gregg's Homepage](https://www.brendangregg.com/?utm_source=chatgpt.com) - His **Systems Performance** book is essentially the field's reference textbook, and his Linux performance talks are excellent. [Brendan Gregg: Overview](https://www.brendangregg.com/overview.html?utm_source=chatgpt.com) - [Brendan Gregg's website](https://www.brendangregg.com/?utm_source=chatgpt.com) 2. **Julia Evans** - Fantastic for developing the *mental models* needed to understand systems. - Particularly good for Linux, networking, debugging, operating systems, and making intimidating low-level concepts approachable. - Her visual explanations are a great complement to Gregg's more rigorous material. 3. **Andi Kleen** - Follow for CPU performance, Linux `perf`, profiling, hardware performance counters, and compiler/runtime performance. - Especially useful once you move from "which process is slow?" to "what is the CPU actually doing?" 4. **Martin Thompson** - Excellent for **hardware/software interaction**: CPU caches, memory, concurrency, latency, mechanical sympathy, and high-performance Java/C++ systems. - Particularly valuable if you're interested in low-latency systems. 5. **Gil Tene** - Follow for **latency, benchmarking, JVM performance, GC, coordinated omission, and measurement methodology**. - His work is especially valuable for understanding why seemingly reasonable benchmarks can produce completely misleading conclusions. 6. **Brendan Gregg + Netflix/USENIX material** - This is worth emphasizing because the **methodology** is more important than memorizing Linux commands. - Gregg's framework covers observability → methodology → benchmarking → profiling → tracing → tuning. [Linux Systems Performance | USENIX](https://www.usenix.org/conference/lisa19/presentation/gregg-linux?utm_source=chatgpt.com) ### Organizations worth following - **USENIX** - particularly **SREcon** and LISA. Their talks contain a lot of serious production-performance work rather than generic "DevOps" content. - **Netflix Technology Blog** - excellent production-scale performance and systems engineering material. Netflix has published tutorials specifically on Linux performance tooling. [Netflix at Velocity 2015: Linux Performance Tools | by Netflix Technology Blog | Netflix TechBlog | Medium](https://medium.com/netflix-techblog/netflix-at-velocity-2015-linux-performance-tools-51964ddb81cf?utm_source=chatgpt.com) - **Linux Foundation** - kernel, eBPF, observability, and low-level Linux ecosystem developments. - **ACM / ACM Queue** - good for deeper systems/performance thinking rather than just tooling. ### If I were learning this from scratch I'd use this sequence: **Brendan Gregg → Linux fundamentals → `perf` → flame graphs → eBPF → CPU/memory architecture → networking → benchmarking → distributed-systems performance.** And I'd strongly recommend learning **methodology before tools**. A performance engineer who knows 50 Linux commands but can't formulate a hypothesis is much less effective than someone who can reason about where latency or resource consumption *must* be coming from. Gregg's own "Linux Systems Performance" talk is a particularly good starting point because it gives you the map of the territory before you dive into individual tools. [Linux Systems Performance | USENIX](https://www.usenix.org/conference/lisa19/presentation/gregg-linux?utm_source=chatgpt.com) If you tell me whether you're aiming at **backend/cloud systems, databases, Linux/kernel, AI/ML infrastructure, or low-latency trading**, I can give you a much more targeted **10–15 person follow list + learning roadmap**.

**Follow experts like Kelsey Hightower, Charity Majors, Cindy Sridharan, and Liz Fong-Jones for deep insights into systems performance engineering and observability. Complement this with web performance voices such as Ilya Grigorik and Patrick Meenan for a broader perspective.** ## 🔑 Key Figures in Systems Performance Engineering Expert| Focus Area| Why Follow ---|---|--- **Kelsey Hightower**| Kubernetes, SRE, reliability| Known for clear explanations of distributed systems and reliability practices. **Charity Majors**| Observability, debugging, resilience| Co-founder of Honeycomb.io, strong advocate for production-first debugging. **Cindy Sridharan**| Systems theory, observability| Writes deeply analytical essays on distributed systems and monitoring. **Liz Fong-Jones**| SRE culture, SLOs, reliability| Former Google SRE, now at Honeycomb.io, focuses on reliability culture. **Ben Sigelman**| Tracing, OpenTelemetry| Co-creator of OpenTelemetry, essential for modern observability stacks. **Bryan Cantrill**| Systems design, hardware/software interplay| CTO of Oxide Computer Company, brings historical and practical depth. **Tammy Butow**| Chaos engineering, incident response| Shares practical lessons from large-scale reliability testing. **Corey Quinn**| AWS cost \+ observability| Adds humor and sharp insights into cloud performance and efficiency. [25 TOP TECH INFLUENCERS WITH PERFORMANCE MONITORING IN 2025](https://www.amraandelma.com/tech-influencers-with-performance-monitoring/)[The Top Engineering Voices to Follow in 2025 - Jellyfish](https://jellyfish.co/blog/the-top-engineering-voices-to-follow-in-2025/) ## 🌐 Web Performance Voices (Complementary) While systems performance engineering often focuses on backend, distributed systems, and observability, **web performance experts** provide crucial insights into frontend and user-facing optimization: * **Ilya Grigorik** – Google engineer, co-chair of W3C Web Performance group. * **Patrick Meenan** – Creator of WebPageTest, Chrome engineer. * **Lara Hogan** – Author of _Designing for Performance_, focuses on frontend optimization. * **Paul Irish** – Works on Chrome DevTools and browser performance. [50+ Web Performance Experts to Follow Online - KeyCDN](https://www.keycdn.com/blog/web-performance-experts) ## ⚠️ Risks & Considerations * **Over-specialization** : Following only one niche (e.g., Kubernetes or frontend) may give a skewed view. Balance backend observability experts with frontend performance voices. * **Tool bias**: Many influencers are tied to specific tools (Prometheus, Grafana, Honeycomb). Their advice may lean toward those ecosystems. * **Rapid evolution**: Performance engineering practices evolve quickly; staying updated requires following multiple voices across platforms (Twitter/X, blogs, podcasts, GitHub). ## 📌 Suggested Approach 1. **Start with core observability leaders** (Charity Majors, Cindy Sridharan, Liz Fong-Jones). 2. **Add practical system builders** (Kelsey Hightower, Bryan Cantrill). 3. **Complement with frontend/web performance experts** (Ilya Grigorik, Patrick Meenan). 4. **Engage with open-source communities** (OpenTelemetry, Prometheus, Grafana) to see real-world discussions. Would you like me to curate a **“starter list” of blogs, podcasts, and Twitter/X handles** so you can follow them directly without needing to hunt each one down?

* **Brendan Gregg** * **Who he is:** Former performance engineering lead at Netflix and Intel; universally recognized authority on OS and cloud performance. * **Why follow:** He authored _Systems Performance: Enterprise and the Cloud_ , popularized the **USE Method** (Utilization, Saturation, Errors), and is a principal educator on Linux internals, performance counters, and eBPF. [Big Thinkers: Brendan Gregg – How Systems Performance](https://build5nines.com/big-thinkers-brendan-gregg-how-systems-performance-engineering-shaped-modern-observability/#:~:text=Gregg%20also%20popularized%20practical%20troubleshooting%20methodologies%2C%20particularly,Method%3A%20*%20Utilization.%20*%20Saturation.%20*%20Errors.) * **Where to find him:** `brendangregg.com`, his textbooks, and his deep-dive conference presentations. * **Bryan Cantrill** * **Who he is:** Co-founder and CTO of Oxide Computer Company; former Sun Microsystems engineer and Joyent CTO. * **Why follow:** A co-creator of DTrace, Cantrill is renowned for his fiercely analytical, highly entertaining talks on kernel debugging, software-hardware interaction, and systems-level failure analysis. * **Where to find him:** Conference keynotes (USENIX, SCALE) and technical podcasts. * **Charity Majors** * **Who he is:** Co-founder and CTO of Honeycomb.io; co-author of _Observability Engineering_ and _Database Reliability Engineering_. [Top engineering voices to follow in 2025 | Blog - Incident.io](https://incident.io/blog/top-engineering-voices-2025#:~:text=Charity%20has%20co%2Dwritten%20Database%20Reliability%20Engineering%20and,work%20cultures.%20Why%20follow%20Charity%20in%202025%3F) * **Why follow:** She focuses on the operational side of performance - how systems behave in production under high load, why traditional metrics fail, and how modern observability bridges performance tracing with software reliability. [Top engineering voices to follow in 2025 | Blog - Incident.io](https://incident.io/blog/top-engineering-voices-2025#:~:text=Cassidy%20is%20the%20perfect%20blend%20of%20technical,%7C%20Honeycomb.%20Charity%20runs%20the%20blog%20charity.) * **Where to find him/her:** Her blog (`charity.wtf`) and LinkedIn/X. [Top engineering voices to follow in 2025 | Blog - Incident.io](https://incident.io/blog/top-engineering-voices-2025#:~:text=Charity%20has%20co%2Dwritten%20Database%20Reliability%20Engineering%20and,work%20cultures.%20Why%20follow%20Charity%20in%202025%3F) * **Adrian Cockcroft** * **Who he is:** Former VP of Cloud Architecture at Netflix and former battery/sustainability technologist at AWS. * **Why follow:** Cockcroft pioneered cloud-native performance engineering, devising the microservices architecture, latency modeling, and cloud tooling strategies that scaled Netflix globally. * **Where to find him:** Industry panels, architectural keynotes, and his technical blog posts on cloud scale. * **Martin Kleppmann** * **Who he is:** Researcher at the University of Cambridge and author of _Designing Data-Intensive Applications_. * **Why follow:** Systems performance frequently bottlenecks at the database and storage layers. Kleppmann’s work breaks down how data structures, replication, consensus algorithms, and networking impact distributed systems performance. * **Where to find him:** `martin.kleppmann.com`, his book, and his academic lectures.

To master systems performance engineering, you should follow industry pioneers who specialize in operating systems, kernel tracing, hardware-software interaction, and large-scale infrastructure observability. Here are the definitive experts, authors, and practitioners you should follow: 1. ** _Brendan Gregg_** * **Why follow:** Widely considered the gold standard authority in systems performance. He is a Partner Kernel Engineer at Intel (formerly at Netflix and Sun Microsystems) and the author of _Systems Performance: Enterprise and the Cloud_ (2nd Edition). His work on BPF, eBPF, flame graphs, and OS performance methodologies is foundational. * **Where to follow:** Read his insights on [Brendan Gregg's Blog](https://www.brendangregg.com/blog/) and check out his tools and methodologies on his [Brendan Gregg's Homepage](https://www.brendangregg.com/). 2. ** _Charity Majors_** * **Why follow:** Co-founder and CTO of Honeycomb, Charity is a massive voice in modern observability, production engineering, and how systems behave under real-world distributed load. Her takes on sociotechnical systems, high-performance engineering culture, and debugging live production systems are invaluable. * **Where to follow:** Follow her commentary and industry insights via [Charity Majors's Blog/X/Socials](https://incident.io/blog/top-engineering-voices-2025). [Top engineering voices to follow in 2025 | Blog - Incident.io](https://incident.io/blog/top-engineering-voices-2025) 3. **Julia Evans** * **Why follow:** Known for her remarkably clear, hand-drawn zines and cheat sheets about low-level systems, Linux internals, networking, and performance troubleshooting tools ( straces t r a c e 𝑠𝑡𝑟𝑎𝑐𝑒 , tcpdumpt c p d u m p 𝑡𝑐𝑝𝑑𝑢𝑚𝑝 , memory profiling). She breaks down intimidating kernel and systems concepts into digestible visuals. * **Where to follow:** Her technical posts and zines are regularly published on her personal site, Julia Evans's Blog. 4. ** _Lorin Hochstein_** * **Why follow:** A leading voice in software reliability, failure analysis, and systems engineering. He frequently writes and speaks about how complex distributed systems fail and how to reason about latency, performance degradation, and feedback loops in large architectures. * **Where to follow:** Track his curated thoughts and engineering leadership insights featured across Incident.io's Engineering Voices. [Incident.io] 5. ** _Dave Taht_** * **Why follow:** A key figure in bufferbloat reduction, network latency minimization, and low-level kernel performance. If your interest leans toward networking stack performance, queue management, and reducing latency jitter at the OS/hardware boundary, his work is essential reading. Would you like me to recommend specific **books and foundational papers** by these authors, or are you looking for resources focused on a particular stack (like **Linux eBPF** , **cloud-native/Kubernetes** , or **database internals**)?