MS Thesis Defense: Extendable Rule-Based Action Generation for Self-Driving Database Systems (Mike Xu)
Database management systems (DBMSs) have become more complex to meet increasingly demanding usage. To owners and operators, the need for a self-driving DBMS that can automatically tune and optimize itself without human intervention is apparent now more than ever. Such a self-driving DBMS considers a set of candidate actions to apply to reach a configuration that improves performance for a given workload. Furthermore, the DBMS would continuously adjust the configuration in anticipation of changing workloads and data distributions. Efforts to... Read More
[15-445/645] Fall 2022 Live Call-in Q&A Lecture
For the final lecture in CMU's Introduction to Database Systems (Fall 2022) course, we are allowing anyone to call in with their database questions. The lecture will be livestreamed via Youtube and you will be able to ask your questions to Prof. Andy Pavlo directly. Livestream: https://youtu.be/MxOKUt6LeeU Audience Call-in: https://cmu.zoom.us/j/99783788428?pwd=R2ZSd2x0SUFnRlNIak5TVk5ubmFjQT09 (Must have Zoom account) Read More
[¡Databases! 2022] CrateDB: Distributed SQL Database Built on Top of Lucene (Marios Trivyzas)
CrateDB is a SQL, open-source, distributed database that makes storage and analysis of massive amounts of data simple and efficient. It offers effective data handling due to a high degree of scalability and availability, real-time query performance, and extensible data models. In this talk, we will discuss the fundamental concepts of CrateDB and highlight what makes it unique in comparison with other databases. In particular, we will cover topics such as dynamic schemas, Lucene index, distributed query execution, logical replication... Read More
[¡Databases! 2022] SplinterDB: A Key-Value Store for Modern Storage Devices (Alex Conway)
We built SplinterDB to address two trends. The first is the modern storage hardware offers significantly higher bandwidth and lower latency. The second is that modern applications at VMware and elsewhere store fine-grain data, such as metadata. Current state-of-the-art key-value stores such as RocksDB fail to fully exploit the capabilities of these devices on these types of workloads. SplinterDB is designed around a novel data structure, the mapped Bε-tree, which combines ideas from the theory of external memory hash tables... Read More
[¡Databases! 2022] TigerBeetle: Magical Memory Tour! (Joran Dirk Greef)
TigerBeetle is an open source distributed financial accounting database designed for mission critical safety and performance to track financial transactions at scale. TigerBeetle is coded to NASA’s Power of Ten Rules for Safety Critical Code. All memory is statically allocated at startup for predictable and efficient resource usage. Function arguments and return values are verified at runtime by over three thousand assertions. Deterministic Simulation Testing accelerates the maturation process of TigerBeetle’s VSR consensus protocol and LSM storage engine, through fault... Read More
[¡Databases! 2022] Gaia: Direct Database Access without Database APIs (Tengiz Kharatishvili)
GAIA database is an unusual engine - it offers a direct access model and supports memory pointers, applications are able to create and navigate complex data structures like graphs with no or very little API overhead: the engine implements transactional memory view with full transactional durability - whatever you do in memory gets captures and persistent in an efficient transparent manner. You get ACID properties of a database but with the look and feel of a "regular" C/C++ code that... Read More
EdgeDB: Replacing SQL and Improving on the Relational Database Model (Michael Sullivan)
EdgeDB is a new database built around an evolution of the relational model that we call "graph-relational". In the graph-relational model, data is represented as strongly typed objects containing set-valued scalar properties and links to other objects. Missing values are represented in the language as empty sets (no NULL!), and have consistent semantics. The query language, EdgeQL, supports convenient fetching and modification of nested data. EdgeQL queries are composable: queries can be used without restriction as subexpressions and bound to... Read More
[¡Databases! 2022] RisingWave: Reinventing(?!) Stream Processing in the Cloud Era (Yingjun Wu)
RisingWave is a cloud-native streaming database. Different from existing streaming systems, RisingWave fully leverages the modern cloud infrastructure to achieve high performance and scalability at a low cost. In this talk, I will walk you through the detailed designs of RisingWave, and discuss how we adopt some old ideas to build a next-generation streaming system. This talk is part of the ¡Databases! – A Database Seminar Series. Zoom Link: https://cmu.zoom.us/j/94466872009 (Passcode 424050) Read More
[¡Databases! 2022] Architecture Insight of OceanBase: A Distributed SQL Database (Charlie Yang)
OceanBase Database, is an open-source, distributed Hybrid Transactional/Real-time Operational Analytics (HTAP) database management system that has set new world records in both the TPC-C and TPC-H benchmark tests. OceanBase Database starts from 2010, and it's built from scratch by Alibaba & Ant Group, it provides features of an enterprise-level database, including the native distributed architecture, financial-grade high availability, transparent horizontal scalability, real-time analytics, multi-tenancy, high cost-efficiency and compatibility with MySQL syntax. OceanBase has been serving all of the critical systems... Read More
[¡Databases! 2022] Neon: Serverless PostgreSQL! (Heikki Linnakangas)
Neon is a Postgres cloud service that separates compute and storage. Separation of compute and storage makes it possible to scale the components independently, and instant startup without a lengthy restore. Amazon Aurora pioneered the separation of compute and storage for OLTP databases. Neon uses the same concept of applying the WAL stream to reconstruct database state in the storage layer. However, Neon takes that concept further, and retains the WAL in a pre-processed format, making it possible to reconstruct... Read More