NoSQL data systems today are everywhere, from social media to retail and scientific applications. For example they are used to store data in graph databases, task queues, stream processing engines, application data caching, event tracking systems, NoSQL stores, and distributed databases. This course is a comprehensive introduction to modern NoSQL data systems. The primary focus of the seminar is on the modern trends that are shaping the industry right now: disk and cache optimized key-value stores, use of compression, tuning based on workload and hardware across the whole memory hierarchy, and adaptivity to read/write ratios. We also study the history of NoSQL data systems, traditional and seminal concepts and ideas such as optimization, indexing, concurrency control, and recovery. In this way, we discuss both how and why NoSQL data systems evolved over the years, as well as how these concepts apply today and how NoSQL data systems will evolve in the future. Our discussion will be based on the latest designs from the research literature and industry as it is adopted in a wide number of modern key-value stores including LevelDB and BigTable at Google, RocksDB at Facebook, Cassandra, HBase and Accumulo at Apache, Voldemort at LinkedIn, Dynamo at Amazon, WiredTiger at MongoDB, and bLSM and cLSM at Yahoo. Relational databases such as MySQL (using MyRocks) and SQLite4 support this design too as a storage engine by mapping primary keys to rows as values.
Who should attend
The seminar is ideal for IT professionals and software engineers who work on any issues related to data management. The seminar covers a broad set of topics to provide insights about how data systems work and how they are designed. In this way, it allows one to make better choices regarding which systems to chose depending on the target data, queries and hardware. In addition, it gives insights to software engineers about best practices and modern system design approaches.
What are the learning outcomes?
• To become familiar with the history and evolution of NoSQL data systems design over the past decade.
• To understand the basic tradeoffs in designing and implementing modern NoSQL data systems.
• To be able to argue about the design of a new NoSQL data system given a data-driven scenario and design a functional prototype.
• To be able to understand which NoSQL data system is a good fit given the needs of an application.
is a professor of Computer Science at Harvard University where he leads DASlab, the Data Systems Laboratory@Harvard SEAS. Stratos works on data system architectures with emphasis on how we can make it easy to design efficient data systems as applications and hardware keep evolving and on how we can make it easy to use these systems even for non-experts. For his doctoral work on Database Cracking, Stratos won the 2011 ACM SIGMOD Jim Gray Doctoral Dissertation award for best thesis world-wide in data management. In 2011 he also won the ERCIM CorBaayen award and was named “most promising young European researcher in computer science and applied mathematics” by the European Research Consortium for Informatics and Mathematics . He is also a recipient of an IBM zEnterpise System Recognition Award, a VLDB Challenges and Visions best paper award and an NSF Career award by the US National Science Foundation. In 2015 he was awarded the IEEE TCDE Early Career Award from the IEEE Technical Committee on Data Engineering for his work on adaptive data systems.
Για περισσότερες πληροφορίες αναφορικά με το σεμινάριο, παρακαλούμε όπως συμβουλευτείτε τα παρακάτω στoιχεία
Επικοινωνία / Δηλώσεις Συμμετοχής:
Tηλ.: 210-5227094, E-mail: firstname.lastname@example.org
Υπεύθυνη Προγράμματος: Ματίνα Τσοχαλή