Multi-version concurrency control (MVCC) is currently the most popular scheme used in modern database management systems (DBMSs). Although the protocol was discovered in the late 1970s, it is used in almost every major relational DBMS released in the last decade. Maintaining multiple versions of data potentially increases parallelism without sacrificing serializability. But scaling MVCC schemes in a multi-core, in-memory DBMS... Read More

The CMU Database Group is holding an orientation meeting for students that are interested in getting involved in research and development of CMU's new flagship database management system (DBMS). Peloton is a high-performance, in-memory relational DBMS for hybrid workloads. The key aspect of Peloton that makes it different from other systems is that it is designed to be completely autonomous... Read More

This is the first meeting of the Database Group for Fall 2016. We will do the usual meet & greet, followed by two talks from PKU/CMU summer interns. Read More

Given a data stream with many attributes and high frequency of events, how to cluster similar events? Can it be done in real time? For example, how to cluster decades of frequent measurements of tens of climatic attributes to aid real time alert systems in forecasting extreme climatic events, such as floods and hurricanes? The task of clustering data with... Read More

In-memory database management systems (DBMSs) outperform disk-oriented systems for on-line transaction processing (OLTP) workloads. But this improved performance is only achievable when the database is smaller than the amount of physical memory available in the system. To overcome this limitation, some in-memory DBMSs can move cold data out of volatile DRAM to secondary storage. Such data appears as if it... Read More

RocksDB is an embedded persistent key-value store for low-latency and high-throughput workload. It has been adapted to a wide range of workloads, including RocksDB as an embedded DBMS and as storage engines of other DBMS systems. Our benchmarks show RocksDB can achieve 126K random reads per second on flash and 7 million random reads per second on memory. RocksDB also... Read More

The IT industry today is undergoing a revolutionary change in how customers deploy and configure their compute resources. Driven by the demand to reduce costs, both in capital and operation expense, these customers are turning to CLOUD or HYBRID-CLOUD solutions. These customers span the spectrum from very small startup businesses to Fortune 500 companies across regions and industries. Oracle Corporation... Read More

The Database In-Memory (DBIM) Option by Oracle is an industry-first dual format in-memory database that maintains transactional consistent data in both row and columnar formats. This unique architecture enables analytic and OLTP workloads to coexist simultaneously, bringing together the best of both worlds. DBIM is the fastest growing database option since its release in 2014, achieving great success with customer... Read More

Using indexes for query execution is crucial for achieving high performance in modern on-line transaction processing databases. For a main-memory database, however, these indexes consume a large fraction of the total memory available and are thus a major source of storage overhead of in-memory databases. To reduce this overhead, we propose using a two-stage index: The first stage ingests all... Read More

Work on theory of distributed systems abstract away from the physical-clock time and use the notion of logical clocks for ordering events in asynchronous distributed systems. Practice of distributed systems, on the other hand, employ loosely synchronized clocks using NTP in a best-effort manner without any guarantees. Recently, we introduced a third option: hybrid clocks. Hybrid clocks combine the best... Read More