Dashboard Insight recently spoke with SQLstream executives about their latest BI offering, as well as thoughts on OLAP, Eigenbase and Firefox 3.5.
Dashboard Insight: Tell us about the history of SQLstream
Damian Black, CEO):
SQLstream was founded in 2003 to simplify the complexities of processing large volumes of dynamic data with a declarative, real-time framework. Our founders include me, a former HP executive for middleware products across Europe, Julian Hyde, author of the Mondrian Relational OLAP engine, and Edan Kabatchnik, a serial entrepreneur and former Oracle executive.
In addition to SQLstream, we've also established Eigenbase, an extensible open source project for the advancement of open relational framework technologies in specialized data management systems. SQLstream is based on the SQL standards-compliant Eigenbase relational engine. Built on open source and open standards, SQLstream 2.0 enables businesses to receive really urgent analytics and simplify data integration.
DI: What areas of the B.I. stack does SQLstream cover?
Damian Black: SQLstream makes possible instant responses to actionable business data and events that impact the bottom line. Based on ISO-standard SQL 2008, SQLstream is a real-time analytics processing and data integration engine that manages data from collection to delivery including, aggregation, validation, cleansing and transformation of data. SQLstream reduces real-time data costs by using continuous ETL and real-time analytics provide real-time results to business intelligence dashboards without expensive database queries and batch processing.
DI: Why is real-time OLAP important?
Julian Hyde, CTO: OLAP (online analytical processing) lets business users ask their own questions as they explore data, and places business information in the context of historical behavior, similar business events happening concurrently, and the behavior of the enterprise as a whole. Real-time OLAP extends OLAP to the most important information: what is happening in the enterprise right now. I'm the author and lead developer of Mondrian, the leading open-source OLAP engine and, as Pentaho Analysis, part of the Pentaho open-source BI suite. I've added fine-grained cache control to Mondrian and integrated it with SQLstream to ensure that Mondrian is always showing the most up to date information, even when querying a data warehouse of many millions of records.
DI: Yes, you've recently partnered with Pentaho and SQL Power to offer a real-time BI solution. Can you tell us more about this new alliance and what it means for your clients?
Damian Black:
Open standards are essential to businesses that want to maximize the value they receive from their solutions while minimizing ongoing costs. As SQLstream and Pentaho are both built on open standards, the partnership came naturally. SQLstream provides real-time analytics for business intelligence while Pentaho provides open source business intelligence solutions. SQL Power provides real-time dashboard services and solutions. As a result of the partnership, customers can now get real-time dashboards and real-time data warehousing from a single, integrated solution.
DI: Can you tell us about the release SQLstream 2.0 and the new features that are offered?
Edan Kabatchnik, COO:
SQLstream 2.0 offers a wide set of new features to enable high volume real-time data processing for real-time dashboarding, data warehousing and data analysis. These new features can be broken into three categories: architecture, core data processing and adapters.
On the architecture side, SQLstream now runs on 64-bit operating systems, allowing real-time data processing on very large memory systems. We've also improved the performance substantially so that it can now process data at 100,000 events per second per server on commodity hardware.
With core data processing, we offer more streaming aggregation. Earlier versions of SQLstream enable aggregation over “rolling” time windows, producing a calculated output row for each calculated input row. These calculations were suitable for Business Event Processing and Real Time Dashboarding applications. SQLstream 2.0 adds support for “tumbling” time windows, producing calculated output rows based on a time-group collections of rows, e.g.. an aggregated summary record every hour. Support for “tumbling” time windows enables SQLstream to support real-time data warehousing and real-time business intelligence applications.
Our adapters now have the capability to stream “upserts” to SQL Server, mySQL, and Postgres, database tables - plus SQLstream 2.0 ensures that Mondrian is always showing the most up-to-date information, even when querying a data warehouse of many millions of records. We've also included a table lookup UDX, web feed adapter, log file reader and agent and more.
DI: Tell us about your support for and use of standards.
Rick Saletta, VP of Marketing: SQLstream is built on open source and open standards. From the founding of Eigenbase to SQLstream 2.0, adherence to the SQL standard is a core value at SQLstream. While other companies may coax customers into vendor lock-in with their own, proprietary, real-time technologies, SQLstream is committed to and supports industry-leading standards including SQL 2008, JDBC, XMI metadata and Eclipse. For the customer, benefits include the ability to create queries, update and reuse real-time applications. These things can be done using either in house SQL talent or readily available - and affordable - SQL consultants and integrators.
Before SQLstream, businesses were required to use low-level middleware which was expensive to develop and costly to maintain. By using standard SQL adapters and queries to create real-time applications, SQLstream reduces integration costs far below the cost of most middleware implementations and even below the cost of SOA (service oriented architecture), forever changing the economics of integration.