August 2016 Summaries

SingleStoreDB has introduced Role-Based Access Control (RBAC) in its latest release, SingleStoreDB Self-Managed 5.1, to protect customer data and provide a critical competitive edge for organizations. This new security feature allows customers to easily scale users and roles without compromising performance, making it suitable for intensive workloads like real-time analytics from the Internet of Things. RBAC was identified as a robust shield against unauthorized access in the 2016 Verizon Data Breach Investigations Report, which cited privilege misuse as one of the most common sources of data breaches. SingleStoreDB's implementation of RBAC aims to maintain flexible integration with various technologies, including Kafka, Spark, and business intelligence software, without exacting a performance penalty. The feature allows organizations in sectors like public sector, financial, and energy companies to address potential data breaches through granular information access rights.

The mobile revolution is transforming our lives with GPS-enabled smartphones, with an estimated 2 billion subscribers expected to grow to 6.1 billion by 2020, providing a rich source of geospatial data that can be used to track movement and extract valuable insights for businesses. SingleStore is designed as a massively parallel scale-out solution for real-time analytics with built-in support for geospatial queries, offering strong transactional semantics and superior performance compared to ElasticSearch in certain use cases, such as updating vehicle geolocations and checking intersection of geolocations with geofences.

The role of Chief Information Officer (CIO) has undergone significant changes over the past thirty-five years, evolving from a senior executive responsible for IT systems to a multifaceted leader who must adapt to rapidly changing business needs. Today's CIO is expected to take on various responsibilities, including managing platforms and systems, investing in security and analytics, identifying trends and threats, ensuring data quality and accessibility, hiring talent, and driving business strategy through the use of cloud infrastructure, big data analytics, and IoT technologies. As IT plays an increasingly important role in core business strategy, CIOs must utilize their knowledge of modern technological trends to identify cost-saving opportunities or create additional revenue streams, thereby maintaining their company's competitive edge.

Gartner's Critical Capabilities Report has recognized SingleStore as the highest scorer in the Operational Data Warehouse Use Case, with a score of 3.77 out of 5, ahead of other notable vendors such as IBM, Teradata, MongoDB and Oracle. This recognition is attributed to SingleStore's strengths in operational BI queries, repetitive queries, continuous data loading, and system availability. The report highlights the growing need for Operational Data Warehouses that can efficiently analyze real-time transaction processing and big data analytics in an in-memory optimized architecture, particularly in industries such as IoT, where vast amounts of data are being generated. The increasing demand for real-time analytics to support business adaptation is driven by megatrends such as mobile computing, risk management, personalization, portfolio tracking, clean energy, and geospatial trends, which require efficient ingestion of continuous data loads to support operational business intelligence queries.

The Berkeley Packet Filter (BPF) tool allows users to run small pieces of code quickly and safely inside the Linux operating system, enabling dynamic tracing of the Linux operating system and expanding its capabilities beyond the operating system to the database. By using Userland Statically Defined Tracepoints (USDT), engineers can ask previously unanswerable questions about how the database interacts with Linux, such as investigating query latencies and lock acquisition times. The power of BPF is demonstrated through the use of a BPF script that traces queries and records their latencies, providing valuable insights into database performance. With increased observability provided by BPF, the SingleStore team can build a transparent and easily accessible modern in-memory database, improving performance and customer interaction with the database.

Tableau has released its latest version, Tableau 10, which introduces several new data-source options to help users gain value from their data. The new features include a SingleStore named connector for real-time analytics and Presto named connector for fast data analysis. Additionally, updates have been made to SQL Server and Oracle databases, including support for contained databases and table functions. There is also increased support for SAP HANA on Mac, including the ability to connect via JDBC and use parameters in SQL statements. Furthermore, Tableau has introduced a new feature called "Remember Connection Details" which saves users' connection settings except for passwords. The latest version of Tableau includes a brand new look and feel and several new features to help users prep, analyze, and share their insights faster.

The latest upsert benchmark showcases SingleStore's performance in handling high-volume updates, achieving 7.9 million upserts per second, which is 6 times faster than Cassandra. This critical use case for internet billing with telcos, ISPs, and CDNs requires fast updates and live dashboards to provide real-time analytics on web traffic, which is expected to grow at a compound annual growth rate of 22 percent from 2015 to 2020. SingleStore's ability to handle upserts efficiently, scale linearly, and run sophisticated queries with ease using SQL makes it an attractive option for businesses in this space. In contrast, Cassandra struggles with large batch sizes and lacks native support for analytics, requiring users to bring in additional components like SparkSQL to support meaningful querying of data.

The text discusses the capabilities of SingleStore, a database solution designed to help businesses operate in today's real-time world by unlocking value from data instantly. The distributed nature of SingleStore allows it to ingest large volumes of data at incredible speed, making it ideal for processing 72TB of data daily or over 6 million transactions per second. Key features include high throughput, query execution architecture, highly scalable, and in-memory and on-disk storage options. Additionally, SingleStore offers Streamliner, an integrated Apache Spark solution that allows users to build real-time data pipelines, providing a unified interface for managing many pipelines, built-in developer tools, fault tolerance, transactional semantics, code reuse, and performance recommendations.

Our performance engineering team has released a new version of dbbench, a widely adopted tool for testing database workloads, with enhanced flexibility and ease of use. The new version includes a tutorial to help new sales engineers get started and a package of high-level tools called dbbench-tools to support advanced workflows. These changes improve the technical proof-of-concept process for customers and open up dbbench for new workloads and use cases, including the ability to display latency histograms and parameterize queries via a `.csv` file. The dbbench-tools package includes tools like dbbench-scaler to help answer complex questions about database scalability, such as how many concurrent users can be handled by a small cluster.