DSLAM

Data Systems Lab At Maryland
Overview
The data systems lab at Maryland conducts research on storage, management, and interaction with data. The group is particularly interested in questions around (1) scalable, concurrent, and distributed data management systems, (2) cloud computing, (3) graph databases (4) data science (5) interactive data-intensive systems, and (6) data exploration.
Projects
Deterministic Database Systems

Key ideas:

  1. Every major commercial database system available today is nondeterministic. This has led to headaches for practitioners concerning database replication, scale-out, and concurrency.
  2. Deterministic database systems make database replication trivial --- the same input is sent to every replica and they are guaranteed not to diverge. By reducing the cost of replication, they facilitate higher replica consistency levels and enable developers to move away from NoSQL systems that are popular as scalable data stores, but rarely have support for strong replica consistency.
  3. Deterministic database systems have shown promise to remove expensive commit protocols in scalable distributed deployments, and enable higher amounts of transactional throughput and concurrency.
For more details see our CACM paper in the August 2018 issue.
Modern main-memory multi-versioned database systems

We are performing research on the design of modern multi-versioned database systems, designed for main-memory deployments. We are exploring an architecture to achieve extremely high throughput, while avoiding the fundamental write skew anomalies that have existed in previous systems affecting both application developers and database users. We are integrating novel techniques for database recoverability and transaction chopping into the multi-versioned database system architecture in order to improve transaction throughput by at least an order of magnitude. We are also investigating techniques for continuous snapshot replication and serving consistent read queries from geo-distributed replicas with bounded staleness.

For more details see our project Website.

 
People

Faculty

Post-docs and Ph.D. students