June 2015 Summaries

The latest release of SingleStore Ops 4.0.31 is now available for download, focusing on providing a smooth on-ramp for developers to get productive with the software. The new features include a Database Speed Test that measures performance, enabling High Availability upgrades in one click, and adding SSL certificates for HTTPS access. Additional functionality includes command line tools for database backup and restore, changing node root passwords, and a trial option for Enterprise Edition directly within Ops. This release is available for both Community and Enterprise Editions of SingleStore.

Tapjoy partnered with SingleStore to implement a database that could power their Mobile Marketing Automation and Monetization Platform, enabling HTAP for real-time ad optimization and overlap analysis, which involves analyzing user interest data quickly and efficiently to serve the right ad to the right person at the right time. This combination of high performance, low latency, and SQL interface enables Tapjoy to process 60,000 queries per second with a response time of less than ten milliseconds. SingleStore's benefits include updates and deletes, interoperability with tools and APIs, and hardware compatibility, allowing customers to build their own white box solution and deploy in public or private clouds. This partnership differentiates Tapjoy from competitors by providing instantaneous access to data up to the last click and concurrent analysis capabilities.

The In-Memory Computing Summit is taking place next week, June 29-30, with various speaking sessions and activities. SingleStore will be at booth #4, offering a brand new t-shirt and providing insights on in-memory computing for peak performance. The summit features sessions on Spark and customer use cases, as well as games and giveaways. Resources such as Gartner Market Guide for In-Memory DBMS and The Modern Database Landscape white paper are available for download, covering topics like real-time analytics, data science platforms, and converged processing.

The SingleStore team's experience at Spark Summit highlighted the benefits of using SingleStore for real-time analytics and hybrid transactional/analytical processing (HTAP) workloads. SingleStore excels in scenarios requiring concurrent ingest and query execution, serving analytics from a real-time operational database, and providing structured database schema and indexes for fast lookups and query execution. The SingleStore Spark Connector enables seamless integration with the Apache Spark framework, allowing users to leverage Spark's expressive programming interface and take advantage of optimized data transfer and computation within SingleStore. By converging transactional and analytical processing in a single database, SingleStore powers tight operational feedback loops that can drive revenue growth and cost savings. The team demonstrated this capabilities by processing real-time energy consumption data from tens of millions of devices and appliances.

Apache Spark is a ubiquitous technology that has gained significant traction in recent times. The platform offers various resources, including videos, blog posts, articles, presentations, and demos, to help users understand its real-world use cases and showcase how it complements other technologies. Pinterest, SingleStore, and Esri are among the notable organizations that have leveraged Spark for their respective applications, such as real-time data pipeline processing and analyzing re-pins across the United States, showcasing the power of Spark in driving business efficiency and revenue. The resources also highlight the shared characteristics of SingleStore and Spark, providing real-world use cases for using these technologies together to achieve real-time analytics. Additionally, the SingleStore Spark Connector offers a comprehensive guide to get started with building real-time applications using Spark and SingleStore, along with performance recommendations and code samples.

Spark Summit showcased MemCity, a simulation measuring energy consumption across 1.4 million households in a futuristic city, using real-time data processing and analytics. MemCity tracks various devices in homes, analyzing data from energy devices measured by the minute, providing insights into energy usage trends and solutions for smarter energy systems. It is built with Apache Kafka, Apache Spark, and SingleStore, solving the problem of ingesting, processing, and serving real-time data across an organization. The simulation runs on four Amazon EC2 nodes, processes 186.67k transactions per second, and enables simultaneous real-time SQL queries. MemCity highlights the importance of smart cities and gathering meaningful insights from sensors to inform urban planning and energy solutions, making long-term planning viable with low-cost hardware.

We're excited to be at Spark Summit in San Francisco next week, where we'll showcase our latest demo highlighting how organizations are using SingleStore and Spark for real-time analytics. Meet us at the event to schedule an in-person meeting or demo, reserve a time slot, and learn about our recent work with customers like Pinterest, who built a real-time data pipeline with SingleStore and Spark Streaming to achieve higher performance event logging, reliable log transport, and faster query execution on real-time data. Additionally, we'll be giving away cool t-shirts and an Estes Drone at booth K6, and you can schedule a meeting or demonstration while there. We also have the SingleStore Spark Connector available for download, which provides optimizations to use Spark and SingleStore together, enabling businesses to build real-time platforms with Apache Spark and extend SingleStore Analytics capabilities.

The concept of "real-time" is subjective and context-dependent, varying across industries and organizations. A real-time system is not just faster, but fast enough to cross a performance threshold where business can reproducibly gain net new value. Two classes of real-time applications exist: machines programmatically making data-driven decisions, such as digital advertising, and humans responding to events and making data-driven decisions in real time, like data center management. The definition of "real-time" is often heuristically defined, allowing stakeholders to establish conventions tailored to their business problems. Real-time systems must be able to process and analyze large amounts of data quickly, with the ability to extract actionable information from it being crucial. Enterprises are deploying real-time data pipelines that can ingest, serve, and query data simultaneously, using technologies like SingleStore, Apache Kafka, and Spark Streaming.

You can quickly evaluate and try out Docker with SingleStore, a distributed database system, using a pre-configured 'cluster-in-a-box' container. With an install of boot2docker.io for Mac or Windows, you can spin up a cluster in just a few minutes. The 'quickstart' container comes installed with SingleStore Ops for management, a single-node cluster, and sample programs referenced in tutorials. If you already have Docker installed, you can jump right in with simple commands. You can create a database, interact with the 'cluster-in-a-box' install of SingleStore, and run benchmarks using Docker commands. For more information, visit the documentation at docs.singlestore.com/latest/setup/docker or try SingleStore in full with a free unlimited scale Community Edition or 30-day Enterprise Trial.