Rohit Ghumare

18 great technical papers from (Meta, Google & Amazon) that helped me crack FAANG+ system design interviews System design interviews, especially for senior roles in FAANG+ companies, test your ability to think about real-world applications, edge cases, and failure scenarios and how to address them. These free research papers & in-depth system design papers are absolute must-reads if you’re preparing for system design interviews: 1. Google File System (GFS):   (https://lnkd.in/gNxqFMwF) 2. MapReduce:  (https://lnkd.in/gc4bqa8W) 3. Borg Cluster Management:    (https://lnkd.in/gUQCuJpD) 4. Amazon DynamoDB:   (https://lnkd.in/ghzF8jRj) 5. Google Chubby Locking Service:   (https://lnkd.in/g-G9hM9J) 6. Spanner:   (https://lnkd.in/giEv9CYm) 7. BigTable:   (https://lnkd.in/gqddJyE8) 8. Facebook Cassandra (Distributed NoSQL DB):     (https://lnkd.in/eD9erCNu) 9. ElasticSearch Architecture:   (https://lnkd.in/gvHAQhas) 10. Redis Architecture:      (https://lnkd.in/gkENeYKh) 11. LinkedIn Kafka (PubSub):    (https://lnkd.in/eGcagdRA) 12. Facebook Memcache (KV Store):   (https://lnkd.in/eYZM5SPb) 13. Dapper Tracing System:   (https://lnkd.in/gmqgEMyW) 14. Zanzibar Authentication System:   ](https://lnkd.in/gJ7m_bha) 15. Pregel Graph Processing:   (https://lnkd.in/g6usKZci) 16. Meta XFaaS (Serverless Functions):      (https://lnkd.in/eHqbPXpH) 17. Napa - Data Warehousing:   (https://lnkd.in/gi34K37Z) 18. Napa - Partitioning Algorithm:   (https://lnkd.in/gjguksXB)

531

26 Comments

This week’s system design refresher: 24 Good Resources to Learn Software Architecture in 2025 SOAP vs REST vs GraphQL vs RPC SQS vs SNS vs EventBridge vs Kinesis Top 5 common ways to improve API performance xAI’s Grok-3 One-Pager SPONSOR US 24 Good Resources to Learn Software Architecture in 2025 The resources can be divided into different types such as: Software Design Books Some books that can help are DDIA, System Design Volume 1 & 2, Clean Architecture, Domain-Driven Design, and Software Architecture: the Hard Parts Tech Blogs and Newsletters Read technical blogs by companies like Netflix, Uber, Meta, and Airbnb. Also, the ByteByteGo newsletter provides insights into software design every week. YouTube Channels and Architectural Resources YouTube channels like MIT Distributed Systems, Goto Conferences, and ByteByteGo can help with software architecture and system design. Azure Architecture Center and AWS Architecture Blog are other important resources. WhitePapers For deeper insights, read whitepapers like Facebook Memcache Scaling, Cassandra, Amazon DynamoDB, Kafka, and Google File System. Software Career Books A Software Architect also needs to develop holistic skills. Books about software career aspects such as Pragmatic Programmer, The Software Architect Elevator, The Software Engineer's Guidebook, and Philosophy of Software Design can help. Over to you: Which other resources will you add to the list?

38

If you're looking to start contributing to a big open-source project like Kubernetes, Argo CD, but couldn't find the right/good-first issues to work on, this is what I would recommend from my experience: Pick any project component (say api-server) and sort issues based on that label. Now assign yourself an issue and start working on it. This ensures that you're understanding the code part by part by picking one specific component at a time. If you're not able to find issues again to work with, go through the codebase and search for 'TODO', these are generally piece of code that needs to get stuff done but have been left knowingly due to certain reasons. Complete those TODOs and thank me later!

64

1 Comment

Scaling for Performance: How Zepto Cut API Latency by 40% with MongoDB Atlas As tech leads, we constantly weigh the trade-offs between performance, scalability, and maintainability. Zepto’s recent shift from a monolithic PostgreSQL-Redis setup to MongoDB Atlas is a great case study in how architecture decisions directly impact business metrics. Key takeaways: 40% latency reduction in critical APIs 6x improvement in traffic handling capacity 14% faster page load times = higher conversions Simplified architecture with microservices and cloud-native design Real-time analytics powered by dedicated analytical nodes The move away from brittle cache layers and toward document-based, scalable systems shows how modern infra choices can drive both speed and growth. Read the full breakdown here: https://lnkd.in/geug3-u3 #ArchitectureMatters #MongoDB #TechLeadership #SystemDesign #Scalability #LatencyReduction #Microservices #CloudEngineering #Zepto #EngineeringWins

42

2 Comments

How does Remix/React Router work under the hood? 👀 I spent 8 hours digging into the source code and took detailed notes to understand its inner workings. If you work with Remix or are curious about its architecture, you'll like it! 👉 https://lnkd.in/euJ8Ji46 And if you're new to React, I've also written about the history of React Server Components, covering fundamental concepts like Virtual DOM and hydration. You can also find it on my blog.

60

3 Comments

This masterclass on stream processing took me about 2+ years of exploring systems design In the post, I will tell you what exactly Is stream processing and how it enables distributed applications to react instantly to events at scale. Here’s a more detailed and actionable breakdown of stream processing based on the transcript and key insights:  1// What is Stream Processing? - Stream processing is the ability to process and act on data in real-time as it flows through a system.  - Traditional systems rely on databases for storing and passing data. Stream processing, on the other hand, reacts to events instantly without waiting for data to be written and read from storage.  - Think of it as a real-time message-passing system between a producer (e.g., sensors or logs) and a consumer (e.g., analytics or recommendation engines).   2// How It Works: The Role of Brokers  - Direct connections between producers and consumers create an O(n²) scaling problem, making systems inefficient.  - Stream processing leverages message brokers (e.g., Kafka, RabbitMQ) to decouple producers and consumers, allowing:     - Efficient communication with O(n) connections.     - Reliability through replication and fault tolerance.     - Scalability for distributed systems.   3// Core Use Cases of Stream Processing  1. Time Windowing:      - Group events into windows like tumbling (non-overlapping) or sliding (overlapping) to analyze activity in real-time.      - Example: Monitoring transactions per minute in fraud detection systems.  2. Change Data Capture (CDC):      - Synchronize a database with derived systems like search indexes or caches in real-time.      - Example: Updating a search engine’s index immediately after a database write. 3. Event Sourcing:      - Save raw events in a message broker, enabling replay or migration to different databases later.      - Example: Migrating to a new database without relying on outdated schema.   4// Making Stream Processing Reliable  - At-Least-Once Processing: Ensure events are delivered reliably even in the face of failures.   - Exactly-Once Processing: Avoid duplicates with:    - Consumer Acknowledgments: The consumer signals back to the broker after successfully processing a message.    - Idempotency: Assign unique keys to messages so re-processing doesn’t cause side effects.   5// Why Stream Processing Matters in Modern Systems  - Real-time fraud detection in finance.   - Instantaneous inventory updates in e-commerce.   - Personalized recommendations in streaming platforms.   - Real-time monitoring for IoT sensors. Stream processing is how you make systems scalable, responsive, and future-proof.

101

2 Comments