Aster Data Systems Delivers Massively Parallel Analytic Database Leveraging Commodity Hardware

Aster demonstrates breakthrough architecture for analytics

Redwood City, Calif. – November 29, 2006 – Aster Data Systems shipped version 1.0 of Aster nCluster that ties hundreds of commodity nodes together to enable large scale data analysis with inexpensive hardware. Aster's software innovations transform off-the-shelf, low-cost hardware into an analytic platform consisting of multiple tiers of intelligent clusters.

Aster nCluster is comprised of Queen nodes for coordination, Worker nodes for distributed analytical processing and Loader nodes for high throughput data loading. The architecture provides network-optimized query execution and POD (Performance Optimized Dimensional) Partitioning™ through a series of patent-pending algorithms that combine global and local executors for maximum parallel processing performance and network efficiency.

“We have put tremendous resources and innovations into our architecture to optimize for network traffic across nodes and enhance query performance,” said Tasso Argyros, CTO and VP of Engineering of Aster Data Systems. “This is really the key area of focus for the future to address scalability and manageability.”

The new platform also addresses high availability by focusing on reducing recovery time. The focus on fast recovery to reduce Mean Time to Recovery (MTTR) is a dramatic departure from traditional approaches that seek to minimize failures requiring expensive hardware (Storage Area Networks, Redundant routers, Error Correcting Code RAM, and Multi-core processors) to increase Mean Time to Failure (MTTF). Version 1.0's automated recovery provides real-time recovery of data and processes across multiple nodes so that a node failure results in continued processing to maintain cluster availability while avoiding data loss.

About Aster Data Systems
Aster Data Systems, Inc. is an early-stage stealth-mode startup company building high-performance, ultra-scalable, affordable data analytics solutions for the enterprise. The company is based in Redwood City, CA and was founded by three Ph.D. students from the Computer Science Department at Stanford University. The Aster database is based on a number of unique, patent-pending innovations.

The Best Insights Possible
Whitepaper: A New Approach for Large-Scale Data Management and Data Analysis
Whitepaper: Deriving Deep Insights from Large Datasets
Research Report: MapReduce and the Data Scientist