Curated Learning Materials for System Architecture and Distributed Computing

The author shares a personal list of valuable learning materials compiled in 2009, focusing on system architecture—a field that blends engineering practice with theoretical research. The list includes classic and influential papers, each accompanied by brief commentary.

1. Operating Systems : Mach microkernel (intro and papers), Exokernel (intro and papers), and Singularity (intro and papers) are highlighted, discussing their motivations, design trade‑offs, and historical impact.

2. Virtual Machines : Notable VMMs such as VMware, Xen, Denali, and Entropia are listed with key publications, emphasizing performance, isolation, and large‑scale deployment considerations.

3. Design Revisited : Papers like “Are Virtual Machine Monitors Microkernels Done Right?” and “Thirty Years Is Long Enough: Getting Beyond C” explore the relationship between microkernels and VMMs and the evolution of programming languages.

4. Programming Model : Works on the drawbacks of threads, the SEDA architecture, and Software Transactional Memory are summarized, illustrating alternative concurrency models.

5. Distributed Algorithms : Foundational papers on logical clocks, the Byzantine Generals Problem, and Paxos are described, with notes on their significance and accessibility.

6. Overlay Networking and P2P DHT : Papers on Resilient Overlay Networks (RON), application‑level multicast, and DHT structures (CAN, Chord, Pastry, Tapestry) are presented, outlining their design goals and practical uses.

7. Distributed Systems : Google’s GFS, MapReduce, BigTable, Chubby, and Sawzall are mentioned, followed by storage‑related papers such as Chain Replication, Dynamo, BitVault, and PacificA.

8. Controversial Computing Models : The article discusses Recovery‑Oriented Computing, Failure‑Oblivious Computing, and treating bugs as allergies, focusing on state preservation and fault tolerance.

9. Debugging : Techniques for black‑box performance debugging and tools like CP‑miner for detecting copy‑paste bugs in OS code are outlined.

The piece concludes with a tribute to Leslie Lamport and a reminder that mastering these topics requires both theoretical study and extensive hands‑on practice.