Storage Systems topics and related papers

In this post, I will distill my own ideas and my own views into a structure for a storage system course. Here, I assume here a 15-weeks course with a single 1 1/2 hour lecture per week (as we have in Germany):

1. Introduction, Overview, Disk Drive Architecture
   Material: Ruemmler, Wilkes An introduction to disk drive modeling
2. Disk Scheduling / SSD
   Material: Iyer, Druschel. Anticipatory scheduling: A disk scheduling framework to overcome deceptive idleness in synchronous I/O, Agrawal et al. Design Tradeoffs for SSD Performance
3. RAID
   Material: Patterson et al. Introduction to Redundant Arrays of Inexpensive Disk (RAID), Corbett. Row-Diagonal Parity for Double Disk Failure Correction
4. Local File Systems
5. Local File System Case Studies: ext3, btrfs
   Material: Valerie Aurora. A short history of btrfs, Card et al. Design and Implementation of the Second Extended Filesystem
6. Local File Structures (Sequential, Hashing, B-Tree)
   Material: Comer. The Ubiquitous B-Tree
7. SAN / NAS / Object-based Storage
   Material: Sacks. Demystifying DAS, SAN, NAS, NAS Gateways, Fibre Channel, and iSCSI
8. Examples: NFS, Ceph, GoogleFS/Hadoop DFS
   Material: Weil. Ceph, A scalable, high-performance distributed file system, Ghemawat et al. The Google File System
9. Snapshots and Log-based Storage Designs
   Material: Brinkmann, Effert. Snapshots and Continuous Data Replication in Cluster Storage Environments, Hitz et al.File System Design for an NFS File Server Appliance, Rosenblum, Ousterhout. The Design and Implementation of a Log-Structured File System
10. Fault Tolerance, Journaling, and Soft Updates
    Material: Prabhakaran et al. Analysis and Evolution of Journaling File Systems, Seltzer et al. Journaling Versus Soft Updates: Asynchronous Meta-data Protection in File Systems
11. Advanced Hashing: Consistent Hashing, Share, and Crush
    Material: Karger et al. Consistent hashing and random trees: distributed caching protocols for relieving hot spots on the World Wide Web, Weil et al. CRUSH: controlled, scalable, decentralized placement of replicated data
12. Caching, Replication
    Material: Nelson et al. Caching in the Sprite network file system, Kistler et al. Disconnected operation in the Coda File System
13. Consistency, Availability, and Partition Tolerance
    Material: DeCandia et al. Dynamo: Amazon’s Highly Available Key-value Store, Helland, Life beyond Distributed Transaction: An Apostate's Opinion
14. Data Deduplication
    Material: Muthitacharoen et al., A Low-bandwidth Network File System, Douglis, Iyengar. Application-specific Delta-encoding via Resemblance Detection
15. Performance Analysis
    Material: Traeger, A nine year study of file system and storage benchmarking (at least parts of it)

As books I would recommend:

• Callaghan. NFS Illustrated
• Pate. UNIX Filesystems: Evolution, Design, and Implementation
• Folk, Zoellick. File structures

For me, a few key points are important:

• To clearly separate between classes of file systems and a concrete example. The best example is the class of network file systems vs. NFS. At the end there should be no much question if something is a inherent property of a class of file systems or of the concrete implementation
• To have enough time to handle the basic concepts independently from concrete usages. For example explaining B-Trees as an important file structures independent from the usage in e.g. BTRFS.
• The concepts are more important than the current technology or standards.