13 top in-memory database products
Forrester Research recently issued its Wave report looking at 13 top products for in-memory databases. Five products fell into the Leader category: IBM dashDB/DB2, Microsoft SQL Server 2016, Oracle Database In-Memory/Times Ten In-Memory Database, SAP HANA and Teradata Database. Six products were named Strong Performers: Aerospike, Couchbase Server, DataStax Enterprise, Red Hat JBoss Data Grid, Redis Labs Enterprise Cluster and VoltDB. Two products were named as Contenders: MemSQL Enterprise and Starcounter.
IBM dashDB/DB2
“IBM DB2 with BLU Acceleration is an in-memory columnar data platform that accelerates real-time and ad hoc analytical and operational workloads for IBM DB2,” Forrester says. “Enterprises using BLU Acceleration often require less data storage because of its data compression technology and the ability to optimize data access patterns without the need of tuning, indexes, and aggregates. IBM dashDB, a managed cloud service, shares its technology base with DB2 BLU, providing in-memory database capabilities in a cloud setting. IBM DB2 on Cloud also provides an in-memory database service in a hosted cloud environment.”
Microsoft SQL Server 2016
“Microsoft offers two in-memory database solutions within the SQL Server platform to address OLTP and OLAP workloads,” Forrester says. “In SQL Server 2016, Microsoft supports an in-memory column store for data warehousing to support faster BI, analytics, and predictive analytics as well as an in-memory OLTP feature to support high-performance transactional applications. Microsoft’s road map focuses on improving its in-memory platform, broader use cases (including the IoT, big data, and real time), faster data ingestion, and supporting real-time analytic queries and global operations.”
Oracle Database In-Memory/Times Ten In-Memory Database
“Oracle offers Oracle Times Ten In-Memory Database and Oracle Database In-Memory products,” Forrester says. “With the combination of these two options, customers can deploy any type of application — transactional, operational, and analytical. Forrester spoke with several customers that run both products to support a platform for multiple workloads. Its road map includes performance enhancements with faster decompression using software-in-silicon, in-memory replication, a high degree of automation, in-memory flash extension, and NVRAM support.”
“SAP HANA is a distributed, shared-nothing, in-memory data platform that supports many use cases, including real-time analytics, translytical, systems of insight, and advanced analytics,” Forrester explains. “Enterprises use the platform for in-memory data marts, SAP Business Warehouse, SAP S/4HANA, and SAP Business Suite. It is also the core component in SAP’s cloud offering, such as SAP Cloud Platform (HCP) and SAP HANA Enterprise Cloud (HEC). SAP HANA Vora extends the platform to support the Apache Spark execution framework to deliver enriched interactive analytics on Hadoop.”
Teradata Database
“Teradata Intelligent Memory (TIM) is integrated into the Teradata Database, which keeps active data in memory for faster and real-time analytics,” Forrester explains. “It uses multi-temperature technology to store and optimize frequently used data in DRAM, warm data in SSD/flash, and cold data on economical disk drives. Teradata’s approach is a hybrid focus, not a dedicated in-memory database. The vendor’s strength lies in delivering high performance and scale for analytical workloads, appliance integration, and strong high availability and disaster recovery as well as broad integration with tools and programming languages.”
“Aerospike is an in-memory database that supports various types of applications, including real-time systems of engagement and digital-business-transformation-driven applications,” Forrester explains. “It is a flash-optimized in-memory NoSQL key-value store database that simplifies the development and deployment of applications needing low-latency access. It currently runs only on Linux and does not support ODBC or JDBC protocols. Aerospike’s road map focuses on cloud, security, API integration, scale, and broader use cases.”
Couchbase Server
“Couchbase Server, an open source NoSQL key-value and document database, offers in-memory technology that appeals to enterprises that need to support higher performance, multi-model, scale, and simplified data access for next-generation applications,” Forrester says. “Firms use Couchbase to support social and mobile applications, content and metadata stores, eCommerce, and online gaming applications. It offers full support for documents, a flexible data model, indexing, full-text search, and MapReduce for real-time analytics.”
DataStax Enterprise
“DataStax is a database platform based on Apache Cassandra, built for performance and availability for the IoT, web, mobile, and extreme-scale applications,” Forrester says. “With more than 400 customers, DataStax continues to execute well to support global applications that demand low-latency access to critical information. Although DataStax cannot run completely in memory, its masterless, shared-nothing, peer-to-peer architecture scales across data centers to deliver a strong, distributed global data platform for applications that demand 24x7 availability and scale.”
Red Hat JBoss Data Grid
“Red Hat JBoss Data Grid comprises a distributed cache and a NoSQL Database technology that delivers a peer-to-peer, no-master/slave architecture designed to scale to hundreds of nodes on commodity hardware,” Forrester explains. “It supports transactional capabilities with configurable full ACID compliance and durability using data replication across servers and data centers. JBoss Data Grid integrates with Apache Spark to support real-time analytics, and it has broad security, offering role-based access control, node authentication, column-masking, and encryption.”
Redis Labs Enterprise Cluster
“Redis Labs offers an open source in-memory database that supports operational, analytical, and operational workloads,” Forrester explains. “It supports both relaxed and strong consistency with a flexible, schema less model that’s built on a shared-nothing symmetric architecture. Today, Redis Cloud runs on AWS, Google Cloud, IBM SoftLayer, and Microsoft Azure and is integrated with several platform-as-a-service environments. Redis Labs has developed additional features and technology that encapsulate the open source software and provide enhanced deployment architecture.”
“VoltDB’s in-memory database combines streaming analytics with transactions in a single integrated platform,” Forrester says. “The product is an ACID-compliant, immediately consistent, and distributed shared-nothing in-memory database. VoltDB relies on horizontal partitioning of data to scale out on commodity hardware, on-premises, or in the cloud. It supports synchronous replication within the database cluster to support high availability. VoltDB is available as open source software under the Affero General Public License (AGPL) as well as under a commercial license.”
MemSQL Enterprise
“MemSQL Enterprise is a distributed in-memory database that delivers full ACID compliance and extreme performance to support transactional, operational, and analytical workloads,” Forrester says. “As a purpose-built database for low-latency data access, MemSQL uses SQL and a horizontally scalable, distributed architecture that runs on commodity hardware or the public cloud. MemSQL’s road map includes further improvement with Apache Spark; enabling further integration with cloud platforms; and improving performance, scale, and administration for larger in-memory deployments.”
“Starcounter, founded in 2006, is a fully ACID-compliant, in-memory NoSQL database that is often used to support high-performance transactional and real-time applications,” Forrester explains. “It combines an in-memory database engine and application server for faster development of high-performance business applications. Customers use Starcounter to build web applications, mobile services, and real-time applications in categories such as gaming, advertising, and telecom. Although Starcounter doesn’t have all of the features found in the Leaders, it offers an economical platform for building next-generation business applications.”