Data Storage Systems - Interaktive Studienpläne der TU Ilmenau

An undergraduate-level understanding of maths, programming, data structures & algorithms, operating systems, and distributed systems is assumed.

After attending the lecture, the students know the principles, methods, and applications of large, distributed data storage systems. They are able to explain important components and principles of data storage systems, such as storage devices, interfaces & protocols, file systems, storage tiering & caching, and deduplication & compression. The students are able to select and apply software tools and methods to analyze and understand the internal processes of data storage systems regarding their functioning and performance. 
Participating in the lecture enables the students to evaluate and assess methods and components of data storage systems. They are qualified to design complete data storage system architectures and implement parts of the data storage system in software based on concrete usage requirements.
As part of the practical exercises, the students can present own solutions to specific tasks, participate in topic-specific discussions, and develop parts of data storage systems alone or in small groups.

1 Introduction (History of storage, difierent kinds of storage, applications;evolution in terms of capacity, performance, and price)

2 Storage Device Hardware and Firmware (Internal organization of storage hardware (e.g., HDDs, SSD); Disk scheduling; SSD FTL components)

3 Protocols & Interfaces (NVMe; SATA, PCIe)

4 Linux I/O Stack (Filesystems layer (incl. VFS); storage device layer)

5 File Systems (Files, directories, and file access methods; disk layout strategies (e.g., inodes, etc.))

6 Benchmarking & I/O Performance Analysis (fio, filebench, basic terms: througput, latency, IOPS; blktrace)

7 I/O Performance Enhancements (Parallel I/O Programming)

8 Replication & Crash Recovery (Mirroring, RAID, Erasure Coding)

9 Storage Tiering & Caching (Storage Hierarchy; Cache eviction strategies)

10 Data Deduplication (difierences to compression, techniques)

11 Distributed & Parallel File Systems (distributed: AFS & NFS, Google FS as scalable example; parallel: Lustre, GFS/GPFS)

12 Key-Value Stores (LSM (e.g., RocksDB); distributed: Amazon Dynamo)

13 Object Storage Systems & Cloud Storage (openStack Swift; S3; Azure Storage; Facebook f4)

14 Recent trends in storage systems & novel storage hardware (Computational Storage; Zoned Storage; SMR disks; NVM; DNA & glass)

Lecture with presentations and blackboard (for face-to-face courses), flipped classroom (recorded videos and seminar-like discussions of the topics via videoconf for virtual teaching), Moodle

This is an incomplete list of various interesting and useful books that will be touched during the course. You need to consult them occasionally.
. Remzi H Arpaci-Dusseau und Andrea C Arpaci-Dusseau, Operating Systems: Three Easy Pieces,
CreateSpace Independent Publishing Platform, 2018.
. Alex Petrov, Database Internals: A deep-dive into how distributed data systems work, O'Reilly, 2019.
. Andrew S Tanenbaum, Modern Operating Systems, Pearson, 2014.
. George Coulouris, Jean Dollimore, Tim Kindberg, Gordon Blair, Distributed Systems: Concepts and
Design, Pearson, 2011