Dream for inspiration

The operational database management system (DBMS; see Note 1) market is defined by relational and nonrelational database management products that are suitable for a broad range of enterprise-level transactional applications. These include purchased business applications such as enterprise resource planning (ERP), customer relationship management and customized transactional systems built by an organization's development team. In addition, we include DBMS products that also support interaction data and observation data (see Note 2) as new transaction types. These products are also used both for purchased business applications, such as ERP, catalog management and security event management, and for customized systems.

Additionally, operational DBMSs should provide interfaces to independent programs and tools that permit, and govern the performance of, a variety of concurrent workload types. There is no presupposition that these DBMSs must support the relational model or the full set of data types in use today. Further, there is no requirement that the DBMS be a closed-source product; commercially supported open-source DBMS products are included in this market.

Operational DBMSs must include functionality to support backup and recovery, and have some form of transaction durability — although the atomicity, consistency, isolation and durability (ACID) model is not a requirement. For open-source DBMSs, maintenance and support must be available from a vendor that owns, or has substantial control over, the source code, and must be offered with a full General Public License (GPL) or an alternative.

For this Magic Quadrant, we define operational DBMSs as systems that support multiple structures and data types, such as XML, text, JavaScript Object Notation (JSON), audio, image and video content. They must include mechanisms to isolate workload resources and control various parameters of end-user access within managed instances of data (see Note 3). Emerging technologies, such as cloud-only DBMSs, are not included; nor are highly specialized engines such as graph-only or object databases, which may perform some transactions for small subsets of operational use cases. Products that "add a layer" to and require or embed a complete or near-complete implementation of another commercially marketed product, such as Oracle MySQL, are not included. Finally, "streaming" engines, whose use cases are dominated by immediate event processing, and which are rarely, if ever, used for subsequent management of the data involved, are also excluded.

For the purposes of this analysis, we treat all of a vendor's products as a set. If a vendor markets more than one DBMS product that can be used as an operational DBMS, we describe them in the section specific to that vendor, but we evaluate all of that vendor's products together as a single entity. Strengths and Cautions relating to a specific offering or offerings are noted in the individual vendor sections. It may be important for organizations to evaluate separate offerings as the range of choices broadens, and as purchasers more frequently pursue best-of-breed strategies.

Operational DBMSs may support multiple delivery models, such as stand-alone DBMS software, certified configurations, cloud (public and private) images or versions, and database appliances (see Note 4). These are discussed and evaluated together in the analysis of each vendor.

Source: Gartner (October 2014)

Vendor Strengths and Cautions

Actian

Headquartered in Redwood City, California, U.S., Actian offers relational DBMS (RDBMS; Ingres) and embedded (PSQL) engines, both suitable for operational use. Following a recent repositioning, Actian focuses primarily on analytical use cases.

Actian did not respond to Gartner's requests for supplementary information or to review early drafts of this section. Our analysis is therefore based on other credible sources, including publicly available information, previous briefings with Actian, and interactions with users of Gartner's client inquiry service.

Strengths

• Large customer base: Actian claimed to have over 210,000 customers at mid-2013. Broad geographic and industry coverage for Ingres, and a loyal following for PSQL, remain in place in 2014.
• Rich portfolio: Actian's offerings provide modern features, including multiversion concurrency control (MVCC), object and geospatial support, and column-level encryption.
• Embeddable offering: Actian's PSQL gives it a means of entering the small-footprint, minimal-administration market that is so important to mobile and Internet of Things applications.

Cautions

• Complex portfolio, focused elsewhere: Actian's recent positioning efforts are aimed at analytics, not operational use cases.
• Deployment challenges: Actian received the lowest scores from customers surveyed in 2013 for ease of use, and very low scores for ease of implementation. Our interactions with users of Gartner's client inquiry service in 2014 paint the same picture.
• Market focus: Gartner's interactions with clients continue to indicate a lack of momentum from Actian in relation to its operational DBMS and support of its Ingres DBMS.

Aerospike

Headquartered in Mountain View, California, U.S., and founded in 2009, Aerospike markets a hybrid in-memory/flash NoSQL DBMS — a real-time data platform — for the operational transaction market. It is available both as an open-source community version and an Enterprise Edition.

Strengths

• Operational DBMS functionality: Aerospike's offering makes hybrid use of DRAM and flash as addressable memory and not as file system support. Synchronous copies and support for multiple data centers add high availability (HA) and disaster recovery (DR) capabilities. Aerospike's hybrid DBMS structure supports JSON and NoSQL key-value data.
• Marketing and hardware ecosystem: Aerospike has strong partnerships with hardware component vendors for DRAM and flash memory. Its marketing focuses on the market segment that requires high transaction rates with near-100% availability.
• Performance: Aerospike's reference customers supported its claims of high performance by awarding it the highest scores for performance of any vendor in this Magic Quadrant. They also gave it the highest score for ease of doing business.

Cautions

• Lack of full functionality: Aerospike is strong in HA functionality but lacking in some basic SQL and NoSQL functions, although it has added some SQL functionality.
• Competitive positioning: With an increasing number of vendors supporting in-memory DBMSs, Aerospike will need to differentiate itself more clearly. Many vendors have much larger marketing programs — a disadvantage for Aerospike.
• HA/DR and ease of programming: Aerospike's reference customers identified difficulties in programming and the management of HA/DR as weaknesses of its product, probably due to the complexities of DR.

Altibase

Headquartered in Seoul, South Korea and Palo Alto, California, U.S., Altibase offers Altibase HDB, an SQL operational DBMS capable of using in-memory, traditional disk or hybrid storage. Altibase XDB is an in-memory-only DBMS.

Strengths

• Performance and support: Reference customers gave Altibase high marks for the overall performance of its operational DBMS, and for its support, professional services and ease of use.
• Broad use case applicability: In addition to applying its technology to unique billing scenarios in the telecommunications sector and real-time flaw detection in manufacturing scenarios, customers use Altibase HDB for analytics and storage of textual and rich-media content.
• Product maturity: Over 85% of Altibase's reference customers reported no problems with its product.

Cautions

• Limited global penetration: Although successful in Asian markets, Altibase has yet to establish much brand awareness or penetration elsewhere. The company has recently established several partnerships with global organizations.
• Shortage of large reference customers: Reference customers themselves stated that Altibase's limited number of globally recognized reference customers made adoption more difficult.
• Narrow product focus: Altibase does not support alternative consistency models or support multimodel capabilities, such as documents and graphs.

Basho Technologies

Newly headquartered in Seattle, Washington, U.S., Basho Technologies offers Riak, a distributed, masterless key-value store. It is available as a free, open-source, Apache-licensed download, as an Enterprise Edition, and as Riak CS, a multitenant cloud object store. Basho offers an Amazon Simple Storage Service (S3) API and a cloud service.

Strengths

• Resilience: Riak provides multi-data-center distribution and replication with automated balancing; it does not fail upon server failure or network partition.
• Rich features: Riak offers secondary indexes, MapReduce, support for JSON, tunable consistency, multiple programming languages, Apache Solr support and pluggable storage engines.
• Growing paid customer base: Basho's customers include one-third of the Fortune 500 companies in North America and EMEA. It also has a strong community, which contributes to the product.

Cautions

• Single architecture focus: Riak's key-value-only architecture limits its broader adoption and therefore restricts Basho to the Niche Players quadrant in the operational DBMS market.
• Growing competition: Major vendors will continue to add key-value functionality (such as Microsoft Azure Tables and Oracle NoSQL, both already available), which will create additional competition for key-value use cases.
• Recent changes to management team and reorganization of company: These suggest that prospective customers should conduct a careful assessment before making major commitments to Basho, even though its funding is strong and likely to grow.

Cloudera

Headquartered in Palo Alto, California, U.S., Cloudera offers Cloudera Enterprise, a commercial version of Apache Hadoop for which Apache HBase provides the operational DBMS capabilities. Cloudera Enterprise is available on-premises, as an appliance and through various cloud providers.

Strengths

• Support for emerging use cases: 80% of Cloudera's reference customers use Cloudera Enterprise for storing and processing machine-generated data, such as clickstreams and sensor data. Using this observational data in transactions is now possible with HBase.
• Scalability: Cloudera's reference customers repeatedly mentioned Cloudera Enterprise's ability to scale to accommodate massive data volumes.
• Stability and ecosystem: Cloudera has raised over $1.2 billion in venture funding and has developed a large partner ecosystem that encompasses every relevant segment of the enterprise software market.

Cautions

• Challenging implementation and use: Reference customers scored Cloudera lowest of all the vendors in this Magic Quadrant for ease of implementation. It also scored poorly for ease of operation and programming, and support and documentation.
• Lack of differentiation: Cloudera's operational DBMS, Apache HBase, is also offered by its competitors.
• Focus: The operational DBMS is only one component of Cloudera Enterprise. It may have to compete for development and support resources with the rest of the product suite.

Clustrix

Headquartered in San Francisco, California, U.S., Clustrix offers a low-administration, shared-nothing, distributed RDBMS, ClustrixDB, with automatic sharding and replication. It is available as on-premises software and in the cloud. Clustrix also provides a managed database as a service.

Strengths

• Performance: Clustrix provides extreme scale-out clustering for performance and availability. Parallel SQL query execution across a cluster supports hybrid transaction/analytical processing (HTAP) use cases.
• Simplicity: The Clustrix database is designed to be largely self-managed, to reduce operational complexity and total cost of ownership (TCO). Integration is simplified by the implementation of the MySQL wire protocol.
• E-commerce focus: Clustrix recently announced a new focus on applying its scale-out DBMS to e-commerce applications that face scaling challenges.

Cautions

• Lack of multimodel capabilities: ClustrixDB offers no support for data types beyond traditional relational ones. More than half its reference customers have deployed another operational DBMS to support nonrelational use cases.
• Value challenges: Clustrix received low scores from its reference customers for overall value for money. However, all reference customers were using the Clustrix appliance, not a software-only product.
• Poor early performance: Half the Magic Quadrant survey respondents who did not select ClustrixDB stated that this was because its product performed poorly during a proof of concept (POC) exercise.

Couchbase

Headquartered in Mountain View, California, U.S., Couchbase offers an open-source, distributed multimodel (document and key value) NoSQL DBMS, Couchbase Server. It is offered in Community, Enterprise and Lite Editions for on-premises, mobile or cloud deployment.

Strengths

• Rich feature set: Couchbase's March 2014 release, Couchbase Server 2.5.1, offers in-memory object caching, automatic partitioning, limited ACID transaction support, "eventual persistence" (optional nonblocking writes to a caching layer), cross-data-center replication, a synchronized, lightweight, embedded JSON DBMS, and MapReduce support. Couchbase also has a strong technology road map.
• Large customer base and revenue growth: Gartner estimates that Couchbase has over 450 customers in several industries. It also claims to have achieved 400% revenue growth in the past year as its deal sizes have increased.
• Financial and partner strength: Couchbase added a $60 million "E" venture funding round in June 2014, which is helping to fund growth in its international presence. Couchbase has established relationships with several hardware partners and leading system integrators.

Cautions

• Quality: The number of responses from Couchbase's reference customers that reported bugs or unreliable software was significantly above the average for this Magic Quadrant.
• Missing functionality: Of the surveyed Couchbase customers, 48% reported some absent or weak functionality. Some of these functions are on the road map, but not implemented yet.
• Competition: Growing pressure from megavendors and MongoDB, especially in document-oriented use cases, is likely as interest in these use cases grows.

DataStax

Headquartered in Santa Clara, California, U.S., DataStax provides DataStax Enterprise, a commercial version of the open-source Apache Cassandra database. The product is downloadable for on-premises operation, as well as through multiple cloud providers.

Strengths

• Customer satisfaction: Reference customers scored DataStax above average for most metrics, and awarded very high scores for overall DBMS performance and their experience of doing business with this vendor.
• Expanding functionality: DataStax has added in-memory transactions, search capabilities, and support for analytics through Apache Spark and Apache Hadoop. Reference customers identified administration and development tools as positives.
• Vibrant community: DataStax has helped develop a robust open-source community around Apache Cassandra through developer and enterprise conferences.

Cautions

• Weak brand traction: Inquiries from Gartner clients mention DataStax only one-third as often as the open-source project Apache Cassandra. The market has yet to consistently associate DataStax with Apache Cassandra.
• Challenging starts: 53% of the respondents who evaluated DataStax did not select it due to poor performance during POC testing. This may indicate a poor fit between the characteristics of DataStax Enterprise and the piloted use cases.
• Quality between versions: Reference customers identified regression bugs when upgrading to new versions. DataStax must continue to invest in improving its quality.

EnterpriseDB

Headquartered in Boston, Massachusetts, U.S., EnterpriseDB supports and markets the PostgreSQL open-source DBMS, which it packages as an open-source community edition and as Postgres Plus Advanced Server, including the Oracle Compatibility Feature.

Strengths

• Community leadership: EnterpriseDB is the primary contributor to the PostgreSQL community. It is responsible for many of the new features of PostgreSQL by contributing to JSON, materialized views and partitioning.
• Functionality: Gartner clients report that the functionality of EnterpriseDB's Postgres Plus Oracle Compatibility Feature is now more than sufficient to run both mission-critical and non-mission-critical applications. Recently, Infor, a major application platform independent software vendor (ISV), added EnterpriseDB as a DBMS platform choice.
• Stability and compatibility: Reference customers commend the compatibility with Oracle, the stability of the DBMS and the product support.

Cautions

• Open-source dilemma: EnterpriseDB must conform to community-led release cycles for its community editions as they go through the open-source process. This can slow the process of enhancing the base open-source product, but not the enterprise edition.
• Market exposure: EnterpriseDB lacks breadth in its sales and marketing operations, which restricts its ability to communicate its message to potential enterprise customers. According to our survey, those that did not choose EnterpriseDB would have been more likely to choose it if they had been more familiar with it.
• Support and documentation: Reference customers reported a lack of local-language support and weak documentation.

FairCom

FairCom, which was founded in 1979, is headquartered in Columbia, Missouri, U.S. and privately owned. FairCom c-treeACE (Advanced Core Engine), one of the oldest NoSQL DBMSs, is a fully ACID, key-value store with both NoSQL (Indexed Sequential Access Method [ISAM]) interfaces and SQL, and supports transactions with an embedded or stand-alone engine.

Strengths

• Strong technology: c-treeACE is a very strong ACID key-value NoSQL DBMS with SQL capabilities and a long history of stability and innovation. Cross-platform support (Unix/Linux/OS X), scalability and strong HA stand out among its capabilities.
• Customer base: FairCom has a large customer base, encompassing both stand-alone and embedded implementations. The OEM (embedded) base is itself large and produces sustainable revenue that enables investment in research and development (approximately 25% of revenue).
• Satisfied customers: FairCom received some of the highest overall scores in our survey, with high marks for customer support, professional services, performance, ease of doing business, ease of operations and HA. Furthermore, over 75% of its reference customers reported no problems with it.

Cautions

• Marketing presence: FairCom lacks presence in the general DBMS market as it has a relatively small marketing budget. Growth largely comes from within its existing customer base.
• Small, largely unknown vendor: FairCom needs greater brand awareness to compete effectively with other operational DBMS vendors, especially the better known NoSQL vendors.
• Pricing: Reference customers identified FairCom's pricing model as an issue. We believe this is because other NoSQL vendors offer an open-source pricing model.

IBM

Headquartered in Armonk, New York, U.S., IBM offers DB2 for z/OS, Linux, Unix, Microsoft Windows and Informix. Depending on the DBMS, IBM offers multiple deployment models, from hardware bundling and appliances to deployment in IBM's SmartCloud or third-party clouds.

Strengths

• Performance and features: Survey participants rated IBM highly (among the top three vendors) for HA/DR and overall performance. In-memory DB2 with Blu Acceleration reflects IBM's early vision for in-memory DBMSs. NoSQL support includes a MongoDB-compatible JSON API for document-style, cloud delivery via its Cloudant acquisition and via Bluemix, and Resource Description Framework (RDF) for graph models.
• Hardware integration: DB2 for z/OS dynamically routes analytics to the IBM DB2 Analytics Accelerator (IDAA), creating an efficient HTAP architecture in a single environment and reducing mainframe MIPS to cut operating charges. Other IBM products, such as IBM PureData System for Transactions, use integrated hardware and software.
• Global presence: IBM provides support, implementation and services in multiple vertical markets. It has one of the IT industry's largest networks of software, hardware and service partners.

Cautions

• Sales execution: Like other megavendors, IBM is opaque in its reporting of revenue and customer numbers, but Gartner's RDBMS market share figures indicate that IBM declined in 2013, losing second place in the market to Microsoft.
• Complexity and pricing: For the second year in a row, survey participants scored IBM low for pricing model suitability. Value for money was considered average. IBM is, however, expanding its aggressive pricing, bundling and simplification efforts.
• Software quality: Surveyed IBM customers rated it worst of all the vendors in this Magic Quadrant for "software with bugs or unreliable software," and below average for ease of implementation and ease of operation.

InterSystems

Headquartered in Cambridge, Massachusetts, U.S., InterSystems was founded in 1978. It markets Caché, which was originally an object-oriented DBMS but is now a hybrid NoSQL/SQL transaction engine. Caché has a strong position in the healthcare sector.

Strengths

• Strong functionality: Caché supports a wide variety of data types with object, NoSQL and SQL models, and has strong replication capabilities for HA/DR (as evidenced by strong scores from its reference customers). Database management is automated, so it requires fewer staff resources.
• Focused execution: After establishing a solid product and a large ISV ecosystem that is embraced by the healthcare industry, InterSystems is addressing other markets and achieving early success. Strong execution in the healthcare sector is one reason why InterSystems has moved into the Leaders quadrant this year.
• Performance: InterSystems received some of the highest scores from reference customers for the overall performance of Caché and for their experience of doing business with the company. This confirms the impression Gartner gets from other interactions with its customers.

Cautions

• Market perception: Although InterSystems has branched out from the healthcare sector, it is still generally perceived as being a healthcare-only provider. It must pursue a stronger market vision to move into the broader operational DBMS market.
• Marketing: InterSystems is a midsize DBMS vendor with potential for continued growth, especially as 60% of its reference customers plan to purchase more from it. Investment in sales and marketing is necessary if InterSystems is to challenge the broader market leaders.
• Pricing: InterSystems received relatively low scores from reference customers for the suitability of its pricing model.

MapR

Headquartered in San Jose, California, MapR provides the MapR Distribution, including Apache Hadoop. The M7 Enterprise Database Edition includes MapR-DB, an operational DBMS compatible with Apache HBase. It is available on-premises and through various cloud providers.

Strengths

• Reliability and performance: Reference customers gave MapR high scores for its HA/DR capabilities and cluster stability. Several commended the performance of MapR's operational DBMS.
• Support for emerging use cases: 90% of MapR's reference customers use M7 to capture and analyze machine-generated data, such as log files, clickstreams and connected device data.
• Differentiation and focus: MapR differentiates itself from similar vendors through its replacement of the Hadoop Distributed File System (HDFS) with the MapR Data Platform, which exposes Network File System access.

Cautions

• Availability of skills: 60% of the survey respondents did not select MapR owing to concerns about the availability of relevant skills within their organization. Additionally, MapR's scores for ease of programming were well below average.
• After-sale support: Reference customers gave MapR lower scores for its support and documentation and for its professional services.
• Complexity: Several reference customers remarked on the complexity inherent in deploying MapR's product and the overall immaturity of its ecosystem.

MariaDB

MariaDB (formerly SkySQL) is headquartered in Espoo, Finland. It markets two products: MariaDB 10, an open-source, in-memory-capable, multimodel RDBMS based on, and fully compatible with, Oracle MySQL; and MariaDB Enterprise, a commercially supported bundle with enterprise-targeted add-on components. Both products are available for Red Hat Enterprise Linux, CentOS Linux, Ubuntu, Debian (which includes MariaDB in its distributions) and Microsoft Windows. The company is headed by the creators of MySQL.

Strengths

• Rich functionality: MariaDB offers multiple storage engines, tunable persistence, ACID support with the InnoDB/XtraDB engine, graph storage with Open Query Graph (OQGraph), and support for Apache Cassandra and JSON.
• Value: In our survey of reference customers, MariaDB received one of the three highest scores for value for money, as it did for suitability of pricing method. It also received one of the highest scores for "no problems encountered."
• Strong community and partner network: MariaDB is at the heart of a vibrant MySQL user community and ecosystem. It partners with Linux distribution vendors, IBM, Fusion-io, and organizations offering products for special-purpose storage engines, management, backup and HA, as well as service providers.

Cautions

• Increased competition: MariaDB is increasingly visible and will face more competition, especially as Oracle's consent decree with the EU regarding MySQL expires in 2015¹ and Oracle becomes more aggressive.
• Scale: MariaDB's reference customers mostly quantified the size of their largest databases as being a few hundred gigabytes at most. To compete at the high end against increasing competition, MariaDB will require more terabyte-size reference customers.
• Fragmented offerings: Several customers remarked on the number of separate pieces in MariaDB's software stack; one noted there are "too many independent tools for managing databases."

MarkLogic

Headquartered in San Carlos, California, U.S., MarkLogic offers a document store DBMS in commercial Essential Enterprise, Global Enterprise and Mobile editions, and free, fully-featured developer versions. Its software can be deployed in VMware and Amazon Web Services environments, and, in collaboration with SGI, as the DataRaptor appliance. In 2014, MarkLogic enters the Leaders quadrant for the first time.

Strengths

• Features: MarkLogic's mature enterprise features are extended with tiered storage, HDFS support, backup to Amazon S3, JSON, mobile replication, full text search, geospatial capabilities, Sparql language support, Resource Description Framework (RDF) import support and a converter for MongoDB. Its road map is rich and ambitious.
• Solid customer base and partner network: We estimate that, following recent sizable wins in the financial services sector, MarkLogic now has over 300 commercial customers. It also has a sizable partner ecosystem, which should add momentum to its recent solid growth.
• Customer relationship: Reference customers gave MarkLogic high marks for their experience of doing business with it.

Cautions

• Pricing challenges: Surveyed customers ranked MarkLogic low in terms of value for money and suitability of pricing model. However, MarkLogic altered its pricing structure in 2013 and lowered its prices, and Gartner expects to see this reflected in future surveys.
• Difficult to use: Of the vendors evaluated in this Magic Quadrant, MarkLogic received the lowest overall score from survey respondents for ease of programming. For continued growth, delivery of planned product enhancements and programming language support is essential.
• Geographic concentration on North America: Well over 80% of MarkLogic's customers are in North America. Its overseas expansion efforts must succeed if it is to compete with global providers.

McObject

Headquartered in Issaquah, Washington, U.S., McObject offers eXtremeDB version 5.0, a small-footprint relational in-memory DBMS with extended array and vector support. Since 2001, millions of copies of eXtremeDB have been deployed worldwide in embedded and real-time applications.

Strengths

• Deployment and configuration choices: Typically embedded, eXtremeDB supports Microsoft Windows, Linux, real-time OSs and the Java Native Interface. Clustered configurations are available.
• Functionality: eXtremeDB provides full ACID and tunable persistence, multiversion concurrency control, 64-bit support and hybrid storage for scalability.
• Partnerships: McObject has partnerships with EMC, Fusion-io, HP, IBM and others. Numerous distributors market its product, and it has customers worldwide.

Cautions

• Marketing: eXtremeDB's multiple engines and horizontal and vertical scalability remain little-known in the market.
• Limited targets: eXtremeDB is still seen as being marketed primarily for embedded applications, although this is changing.
• Customer satisfaction: Surveyed customers gave McObject low scores for HA/DR, ease of implementation and ease of operation. These are not all new issues. Given the wide distribution of eXtremeDB version 5.0, they reflect a failure to address continuing challenges.

Microsoft

Headquartered in Redmond, Washington, U.S., Microsoft markets its SQL Server DBMS for the operational DBMS market, as well as Microsoft Azure SQL Database (a database platform as a service) and Microsoft Azure Tables. Microsoft now has in-memory row-store technology for transactions in SQL Server 2014.

Strengths

• Market vision: Microsoft's market-leading vision consists of in-memory computing (SQL Server 2014 now has full transaction in-memory support), NoSQL (with a new document-store DBMS), cloud offerings (both cloud-only and hybrid cloud), use of analytics in transactions (HTAP) and support of mobility. Its vision for in-memory computing and putting the "cloud first" is ahead of its competitors.
• Strong execution: Microsoft SQL Server is an enterprisewide, mission-critical DBMS capable of competing with products from the other large DBMS vendors. Gartner's 2013 market share data shows Microsoft taking second place from IBM in terms of total DBMS revenue.
• Performance and support: Reference customers were very positive, with the performance of SQL Server, documentation, support, ease of installation and operation all rated highly. Only 7% reported problems with the DBMS overall.

Cautions

• Lack of an appliance: Microsoft still lacks an appliance for transactions (one comparable to its SQL Server Parallel Data Warehouse appliance), whereas its major competitors (IBM, Oracle and SAP) all offer one.
• Market image: Although SQL Server is an enterprise-class DBMS, Microsoft continues to struggle to dispel a perception of weakness in this area. Inquiries from Gartner clients demonstrate a continuing perception that SQL Server is not used for mission-critical enterprisewide applications — a view that inhibits wider use of SQL Server as a primary, enterprise-class DBMS.
• HA/DR and pricing issues: Reference customers again found the pricing model for SQL Server unacceptable (they gave it the lowest overall rating of any vendor in this Magic Quadrant) and blamed the price changes that came with SQL Server 2012. Microsoft also received one of the lowest overall scores for ease of implementing HA/DR.

MongoDB

Headquartered in New York City, New York and Palo Alto, California, U.S., MongoDB offers an open-source, document-style DBMS, as well as MongoDB Enterprise, a commercial offering available in various service tiers. MongoDB Enterprise is available through several cloud providers, as well as on-premises.

Strengths

• Customer satisfaction: MongoDB received high scores for every measurement of customer satisfaction in the reference customer survey.
• Improving enterprise capabilities: Recently announced partnerships with analytics and data integration vendors enable MongoDB to tell a well-rounded information management story.
• Operational and management support: Continued investment in the MongoDB Management Service simplifies the running of a large cluster in terms of monitoring, backup and recovery, and provisioning.

Cautions

• Increasingly competitive landscape: Over the past year, several vendors have introduced features that compete with MongoDB's core value proposition. MongoDB will face more pressure to differentiate its offerings against entrenched competitors.
• Growing pains: Although many reference customers reported that MongoDB is easy to get started with, several reported challenges architecting MongoDB for large-scale deployments.
• Trendiness with developers: MongoDB's popularity among developers means it is often selected before application requirements are understood. This can result in a poor fit of DBMS capabilities to the application.

Neo Technology

Neo Technology is headquartered in San Mateo, California, U.S. Neo4j is a native graph-style NoSQL DBMS capable of handling transactions with ACID support and clustering for scalability and HA. Neo became an incorporated company in 2011, but began developing its product much earlier, in 2000. Neo4j, which was first used in a production environment in 2003, is offered as both an open-source version and an Enterprise Edition.

Strengths

• Native graph DBMS: Neo4j is a native graph-style DBMS (as opposed to an existing DBMS to which graph capabilities have been added). It is engineered for performance with transactional ACID capabilities in a single instance and offers tunable consistency across clusters for scalability.
• Growth: Since its founding, Neo has seen strong growth from both its open-source version and Enterprise Edition.
• Performance and ease of use: Reference customers identified performance, ease of operation and implementation, and ease of doing business as strengths of Neo.

Cautions

• Graph model: The graph DBMS model is difficult to understand, which lengthens the learning curve. This problem is exacerbated by growing hype from other vendors about the introduction of a graph model on a nongraph DBMS model.
• Small vendor: Although the Neo4j product is over 10 years old and has grown consistently, Neo remains a small vendor that faces all the issues of risk and stability associated with small vendors.
• Pricing model and HA/DR: Reference customers identified Neo's pricing model as an issue. They also expressed dissatisfaction with the product's HA/DR capabilities.

NuoDB

Headquartered in Cambridge, Massachusetts, U.S., NuoDB provides an operational DBMS designed to scale horizontally and elastically. In addition to being available in on-premises and developer editions, NuoDB's product is available on Amazon Web Services.

Strengths

• Delivery and support: NuoDB received top scores for its support and documentation, professional services and ease of programming. We believe this is enabling it to win contracts to replace other vendors in several locations.
• Rapid deployment: On average, reference customers estimated it took less than five months to deploy NuoDB's product in production environments.
• Support for emerging use cases: 80% of NuoDB's reference customers use it for capturing and analyzing machine-generated data, such as clickstreams, log files and connected device data.

Cautions

• Inconsistent experience: Although NuoDB received several top scores for service delivery, documentation and support, reference customers that weren't full of praise were highly critical. As a new vendor, NuoDB is still perfecting its service and support.
• Slow momentum: NuoDB has not established a footprint in the developer community, which commonly provides informal support and initiates tool development. Gartner clients have yet to show interest in NuoDB during calls to our inquiry service.
• Nascent partner ecosystem: NuoDB's partner program is still developing and has yet to attract the necessary numbers to supplement the company's sales and implementation efforts.

Oracle

Headquartered in Redwood Shores, California, U.S., Oracle markets a complete set of DBMS products for operational systems. These include Oracle Database, Oracle TimesTen, Oracle Berkeley DB, Oracle NoSQL Database and MySQL. In addition to software, several of Oracle's DBMSs are available in engineered systems (appliances).

Strengths

• Broad range of offerings: Oracle has the broadest product portfolio in the market, covering different DBMSs for multiple purposes (RDBMS, NoSQL, streaming data and mobile). Also, it offers delivery in the cloud, on appliances and as stand-alone software. According to Gartner's 2013 market share numbers, Oracle remains in first place for total DBMS revenue market share.
• Functionality: Oracle offers extensive functionality, with many new features (such as the JSON data type and Temporal, which replaces Total Recall) and options such as the Oracle Database In-Memory and Oracle Multitenant options, the latter moving multitenancy to the DBMS layer and reducing support and maintenance. Oracle is also pioneering DBMS functionality on silicon, with new SPARC M7 and T7 chips scheduled for delivery in 2015.
• Solid performance and availability: Reference customers again identified Oracle's DBMS performance and availability as primary reasons for implementation.

Cautions

• Public perception of vision: Oracle's marketing continues to downplay its responses to market trends (such as in-memory functionality) until it announces products. Oracle customers who use Gartner's client inquiry service continue to show confusion and disillusionment in this regard, as they have to make assumptions about Oracle's road maps and vision.
• "Push back" on appliances: Users of Gartner's client inquiry service show a reluctance to purchase products (such as engineered systems) due to perceived "lock in" to Oracle's proprietary systems — some functions, for example, are available only on Oracle hardware and appliances, such as those in Exadata Storage Server software.
• Low cost/value and bugs: Reference customers consider Oracle's products to be expensive and therefore that they have the lowest value proposition. Oracle also received one of the highest scores for bugs reported. Finally, in recent Gartner surveys,² Oracle received the lowest score for ease of doing business.

Pivotal

Pivotal, a spinoff from EMC, is headquartered in San Francisco, California, U.S. It released Pivotal GemFire XD in April 2014, combining GemFire, its distributed in-memory data grid, and SQLFire, its distributed, memory-optimized SQL database, with Pivotal HD, its Hadoop distribution incorporating Hawq (based on the Greenplum massively parallel processing [MPP] column-store DBMS). It is available with Pivotal CF for cloud-based deployment.

Strengths

• Rich functionality: By combining an in-memory transactional engine with Pivotal HD's Hawq analytic SQL engine, Pivotal enables large HTAP-style combinations of real-time transaction and event processing for closed-loop systems. It offers HA, active-active deployment and rolling upgrade support.
• Flexible usages: GemFire provides native object-oriented and REST interfaces; Hawq provides SQL analytics. GemFire XD supports structured data, geospatial data, objects, JSON and key-value pairs.
• Strong global organization: Pivotal has the resources and installed base of EMC as key assets. They include manufacturing, research, and presale and postsale support worldwide.

Cautions

• Maturity: Integrating multiple products is a complex process, and GemFireXD has been on the market for only a few months. Pivotal received the lowest survey scores for support and documentation, and below-average scores for ease of implementation and ease of operation — though this is not unusual for an early-stage product. Its analytic appliance does not yet have an operational counterpart.
• Pricing and pricing model: Pivotal received very low survey scores for value for money and suitability of pricing method, though its new simplified and flexible pricing model should help to improve matters. Pivotal is one of the world's biggest startups, and its revenue will need to grow rapidly to justify EMC's investment.
• Market awareness: The number of inquiries received by Gartner regarding Pivotal products fell to nearly zero after Pivotal was spun off, and has only recently begun to recover. Pivotal's use of the EMC and VMware sales organization is starting to be felt, which is a positive sign.

SAP

Headquartered in Walldorf, Germany, SAP has several DBMS products that are used for transaction systems: SAP Adaptive Server Enterprise (ASE), SAP SQL Anywhere and SAP Hana. Both SAP ASE and SAP SQL Anywhere are available as software only, while SAP Hana is marketed as an appliance.

Strengths

• Leading vision: SAP remains a leader with its vision for HTAP: It now supports most of the SAP applications that run on Hana on a single in-memory database used for transactions and analytics. SAP reports that over 1,000 customers have purchased part or all of SAP Suite on Hana in just over one year of general availability, which underlines the market's interest in HTAP.
• Strong DBMS offerings: SAP has seen strong growth in SAP Hana, SAP ASE (now certified for SAP applications) is growing strongly, and SAP SQL Anywhere continues to lead the mobility market in terms of functionality.
• Performance: Reference customers again identified performance (scalability and reliability) as a major strength for SAP, awarding it one of the highest scores. Additionally, SAP received the highest score for ease of operation across its DBMS products.

Cautions

• Marketing communications: Interactions with users of Gartner's client inquiry service confirm confusion over SAP's messages about how its DBMS products integrate, where each product can be used, what SAP Hana can and cannot do, and most importantly, whether SAP Hana will be required in the future.
• Lack of skills: As inquiries from Gartner clients make clear, skills to support SAP Hana remain scarce in the market.
• HA/DR problems: SAP's reference customers (and especially users of SAP Hana) reported the lowest level of satisfaction with their vendor's HA/DR capabilities. For the second year, SAP also received the lowest score for clients' experience of doing business with it; similarly, in recent Gartner conference surveys,² SAP received the second-lowest score for ease of doing business.

TmaxData

Headquartered in Bundang-gu, South Korea, TmaxData provides Tibero, an SQL RDBMS featuring various clustering options, integrated encryption and compatibility with other vendors' DBMS products.

Strengths

• Satisfied customers: Reference customers scored TmaxData above average on most satisfaction measures, and substantially above average for ease of implementation.
• Support for mixed workloads: Tibero Active Cluster (TAC) is aimed at transactional workloads using shared disk clustering, while Tibero InfiniData operates in a shared-nothing environment and integrates with Hadoop for analytics workloads.
• Several pricing options: Tibero is available in three editions. Each offers different core/processor pricing options and features.

Cautions

• Limited geographic traction: Despite opening offices in several countries in 2013 and 2014, TmaxData has yet to gain significant traction outside South Korea.
• Uneven postsale support: Reference customers gave TmaxData low scores for support and documentation, and for professional services.
• Software quality: Although no problems were reported consistently, only half of TmaxData's reference customers reported encountering no problems with its products.

VoltDB

Headquartered in Boston, Massachusetts, U.S., VoltDB markets an in-memory row-store operational RDBMS that is increasingly available via vertical-market partners. VoltDB version 4.3, released in May 2014, is an open-source DBMS available as software only.

Strengths

• Technology and vision: Substantial SQL-92 functionality, in-memory DBMS architecture and precompiled Java stored procedures drive VoltDB's high performance in support of HTAP use cases. Tunable consistency and JSON support have been added to give developers more flexibility.
• Customer satisfaction: VoltDB, the only open-source in-memory DBMS vendor, received above-average scores from reference customers for suitability of pricing and professional services.
• Performance and value: As expected for an in-memory DBMS vendor, VoltDB received high scores from reference customers for overall performance of the product and value for the price paid.

Cautions

• Competitive challenges: VoltDB's small, U.S.-centric sales organization and modest ecosystem are growing, but their small size still limits its ability to reach new customers.
• Feature gaps: Although VoltDB customers are overwhelmingly on the newest release, its reference customers identified "some absent or weak functionality."
• Revenue model: VoltDB's revenue remains relatively modest according to Gartner's estimates. Extra funding will be needed to achieve the needed growth. A recent $8 million of Series "B" round venture capital funding should help.

Vendors Added and Dropped

We review and adjust our inclusion criteria for Magic Quadrants and MarketScopes as markets change. As a result of these adjustments, the mix of vendors in any Magic Quadrant or MarketScope may change over time. A vendor's appearance in a Magic Quadrant or MarketScope one year and not the next does not necessarily indicate that we have changed our opinion of that vendor. It may be a reflection of a change in the market and, therefore, changed evaluation criteria, or of a change of focus by that vendor.

Added

• Cloudera: This Hadoop distribution vendor now supports operational transactions through the use of Apache HBase.
• FairCom: This vendor sells a NoSQL DBMS.
• MapR: This Hadoop distribution vendor now supports operational transactions through the use of Apache HBase.
• MariaDB: This vendor offers a MySQL-compatible, open-source RDBMS.
• Neo Technology: This vendor's graph DBMS engine supports transactions.
• Pivotal: This vendor's platform has an in-memory DBMS to support transactions across multiple sources of data.
• TmaxData: This South Korean vendor, which also competes in North America, offers an RDBMS.

Dropped

• Orient Technologies: This vendor failed to meet the inclusion criteria.

To be included in this Magic Quadrant, vendors and products had to meet the following criteria.

Software availability: Vendors had to have DBMS software generally available for licensing or supported download for at least a year, as of 1 July 2014. Products that have been commercially available for over 10 years but have not grown in terms of revenue at or near the market rate for several years are excluded.

Software releases: We use the most recent generally available release of the software to evaluate current technical capabilities. We do not consider beta, "early access," "technology preview," "ramp up" or other releases that are not generally available. As regards reference customers and their survey responses, all versions currently used in production are considered. When older versions are in use, we consider whether later releases may have addressed any reported issues, but also the rate at which customers move to newer versions.

Feature availability: Product evaluations include technical capabilities, features and functions present in the product or supported for download through midnight, U.S. Eastern Daylight Time on 1 July 2014. Capabilities, product features and functions released after this date could be included at Gartner's discretion and in a manner Gartner that deemed appropriate to ensure the quality of this publication. We also considered how such later releases could affect the end-user experience.

Customers and revenue: Vendors had to generate a minimum of $20 million U.S. dollars in verifiable annual software revenue or have at least 100 verifiable and distinct customer organizations with operational DBMSs in production. In addition, each vendor had to identify a minimum of 10 reference customers who would respond to Gartner's approved reference customer survey (for this year's Magic Quadrant, the survey questionnaire was in English only). Revenue can be from licenses, support and/or maintenance. Gartner may include additional vendors based on undisclosed references in cases of known use for classified but unspecified use cases.

Support: The vendor had to provide support for its operational DBMS product(s). We also considered products that control or participate in the engineering of open-source DBMSs and their support. We required that a DBMS meet Gartner's definition of a DBMS (see Note 1).

Services: Vendors participating in the operational DBMS market had to demonstrate their ability to deliver the necessary services to support transaction systems via the establishment and delivery of support processes, professional services and/or committed resources and budget.

Geographical availability: Vendors had to demonstrate support for operational DBMS customers in at least two of the following major regions: North America, Latin America, Europe, the Middle East and Africa, and Asia/Pacific.

Excluded products: Product categories excluded from this Magic Quadrant (see also Market Definition/Description) are:

• Embedded-only DBMS products
• Data warehouse-only DBMS products
• DBMS products available only as a cloud service
• Prerelational DBMS products
• Graph-only DBMS products
• Data grid products
• Complex-event processing or streaming-data engines

Ability to Execute

The Ability to Execute criteria are primarily concerned with vendors' capabilities and maturity. Criteria under this heading also consider products' portability, ability to scale, and ability to run in different operating environments (giving the customer a range of options). Ability to Execute criteria are crucial to customers' satisfaction and success with a product, so reference customer interviews and survey responses are weighted heavily throughout.

Product or service includes the technical attributes of the DBMS(s), as well as features and functions built specifically to manage the DBMS when used as a platform for transactions, interactions and observations. We include HA/DR, performance and scalability, support for multiple deployment options (such as virtualization, cloud and hybrid cloud/on-premises), and support for multiple programming languages and new hardware and memory models. These attributes are evaluated across a variety of database sizes and application workloads. We also consider the automated management, tools and resources necessary to manage a database environment, especially as it scales to more complex application workloads. Finally, we consider the flexibility of the DBMS to incorporate new data types, application types and new requirements for distributing data across multiple servers and geographies.

Overall viability includes corporate aspects such as the skills of personnel, financial stability, investment in research and development, and mergers and acquisitions. It also covers the management's ability to respond to market changes and the company's ability to weather market difficulties (crucial for long-term survival). Vendors are further evaluated on their ability to establish dominance in meeting specific market demands.

Sales execution/pricing covers the price/performance and pricing models of the DBMS products, and the ability of the sales force to manage accounts (based on feedback from interviews, surveys and interactions with users of Gartner's client inquiry service). We also consider the market share of the DBMS products. Also considered are the diversity and innovativeness of packaging and pricing models, including the ability to promote, sell and support the products globally.

Market responsiveness/track record includes the diversity of the vendor's offerings in response to changing market demands — for example, its ability and flexibility to offer appliances, cloud deployment, new data types and new programming requirements. We consider general market perceptions of the vendor and its products. We assess both the vendor's ability to adapt to market changes during the previous 18 months and its flexibility in response to market dynamics over a longer period.

Marketing execution evaluates such activities as lead generation, including traditional methods and Internet-enabled trial software delivery, and the execution of channel development through partnering agreements (including coseller, comarketing and colead management arrangements). Also considered are the vendor's coordination and delivery of education and marketing events throughout the world and across vertical markets, and the creation and support of "community" activities that help to raise awareness and develop skills among buyers and prospective customers.

Customer experience evaluations are based primarily on reference customer surveys and interviews conducted for this report, as well as discussions with users of Gartner's inquiry service during the previous six quarters. We consider the vendor's track record of POCs, customers' perceptions of its product(s), and customers' loyalty to the vendor (this reflects their tolerance of its practices and can indicate their level of satisfaction). Additionally, customer input regarding the applicability of products to limited use cases can be significant, depending on the success or failure of the vendor's approach in the market.

Operations covers the alignment of the vendor's organization, as well as whether and how this enhances its ability to deliver. Aspects considered include field delivery of appliances, manufacturing (including the identification of diverse geographic cost advantages), internationalization of the product (in light of both technical and legal requirements) and adequate staffing.

Source: Gartner (October 2014)

Completeness of Vision

Completeness of Vision encompasses a vendor's abilities to understand the functional capabilities needed to support operational environments, to develop a product strategy that meets the market's requirements, to comprehend overall market trends, and to influence or lead the market when necessary. A visionary leadership role is necessary for the long-term viability of both product and company. A vendor's vision may be demonstrated — and improved — by its willingness to extend its influence throughout the market by working with independent third-party application software vendors that deliver both additional functionality for the operational environment and commercial off-the-shelf software. A successful vendor will be able not only to understand the competitive landscape of operational transactions but also to shape the future of this field.

Market understanding assesses a vendor's ability to understand the market and shape its growth and vision. In addition to examining a vendor's core competencies in this market, we considered its awareness of new trends, such as the increasing sophistication of end users, growing scalability needs (especially across server clusters), the cloud as a platform for DBMSs, the demand for in-memory computing and HTAP, use of new consistency models, and the growing desire to use data structures other than relational ones.

Marketing strategy refers to a vendor's marketing themes, research-and-development focus, and ability to choose appropriate target markets and third-party software vendor partnerships to enhance the marketability of its products. For example, we considered whether the vendor encourages and supports ISVs in its efforts to support the DBMS in native mode (via, for instance, comarketing or coadvertising with "value added" partners). This criterion includes the vendor's responses to the market trends identified above and any offers of alternative solutions in its marketing materials and plans.

Sales strategy assesses how a vendor designs and targets its channels and partnerships developed to assist with selling. This is especially important for younger organizations, as a good sales strategy can enable them to greatly increase their market presence, while maintaining lower sales costs (for example, through free downloadable community editions, coselling and joint advertising). This criterion also covers a vendor's strategy for communicating its vision to its field organization and, therefore, to clients and prospective customers. Also included are pricing innovations and strategies, such as new licensing arrangements and cloud-based models for elastic provisioning to support peak demand.

Offering (product) strategy covers the design of product packaging and deployment options, including the availability of cloud versions, developer editions and appliances based on the vendor's DBMS. Vendors should demonstrate a diverse strategy that enables customers to choose what they need to build a complete solution for an operational environment. Also covered are partners' offerings that include technical, marketing, sales and support integration.

Business model covers how a vendor's model of a target market combines with its products and pricing, and whether the vendor can generate profits with this model, judging from its packaging and offerings. Additionally, we consider reviews of publicly announced earnings and forward-looking statements relating to an intended market focus. For private companies and to supplement publicly available information, we use proxies for earnings and new customer growth, such as the number of Gartner clients indicating interest in, or awareness of, a vendor's products during calls to our inquiry service.

Vertical/industry strategy affects a vendor's ability to understand its clients. Aspects such as vertical-market sales teams and partnerships with vertical-market service providers are considered.

Innovation assesses vendors' approach to developing new functionality that aligns with its market and offering strategies by allocating and managing research-and-development spending and leading the market in new directions. Uses of new storage and hardware models are key examples of this.

Geographic strategy, including a vendor's worldwide reach, is assessed by considering its plan to use its own resources in different regions, as well as those of subsidiaries and partners. This criterion considers a vendor's plan for supporting clients throughout the world, around the clock, and in many languages. Anticipation of regional and global economic conditions is also considered.

Source: Gartner (October 2014)

Quadrant Descriptions

Leaders

Leaders generally demonstrate the most support for a broad range of operational applications, based on support for a wide range of data types and large numbers of concurrent users. These vendors demonstrate consistent customer satisfaction and strong customer support. Many have competed in this market for many years, and have built a wide partner ecosystem for their products. Hence, Leaders generally represent the lowest risk for customers in the areas of performance, scalability, reliability and support. As the market's demands change, so Leaders demonstrate strong vision in support not only of the market's current needs but also of emerging trends. Finally, the messaging, product research and development, and delivery of leaders are in line with today's market and with new trends in both DBMS software and hardware technology.

Challengers

Challengers are stable vendors with strong, established offerings but a relative lack of vision. It is normal for some vendors to have high scores for execution but to lag in terms of the adoption levels and vision needed for leadership. Challengers normally show strong corporate viability and financial stability, and demonstrate strong customer support. However, they lack the vision to support some of the new trends in the operational DBMS market, such as support for interaction and observation data in transactions, or a road map for moving toward in-memory DBMS capabilities. Although they may be lacking in relation to some of the market's innovative concepts, Challengers offer stability, simplicity of installation and support, and strong performance. As with the Niche Players, Gartner considers support of a limited number of data types and hardware models as evidence of limited vision.

Visionaries

Visionaries take a forward-thinking approach to managing the hardware, software and end-user aspects of an operational DBMS environment. Visionaries typically have innovative ideas for new functionality and advanced use of new hardware. They have the requisite number of production customers, but lack the market momentum of Leaders. In this market, Niche Players are often young, small and innovative vendors with great new ideas that are spurring on the more mature vendors and the market in general.

Niche Players

Niche Players generally deliver a highly specialized product with limited market appeal. Frequently, a Niche Player provides an exceptional operational DBMS product, but is isolated or limited to a specific end-user community, region or industry. Although the solution itself may not have limitations, adoption is limited. Niche Players contains vendors from several categories:

• Those with an operational DBMS product that lacks a strong or a large customer base
• Those with an operational DBMS that lacks the breadth of functionality of those of Leaders
• Those with new operational DBMS products that lack general customer acceptance or the proven functionality to move beyond niche status

Niche Players typically offer smaller, specialized solutions that are used for specific operational and transactional applications, depending on the client's needs.

At one time, Gartner viewed the online transaction processing (OLTP) DBMS market as very mature, with few new entrants to challenge the status quo. However, in recent years, the market has changed rapidly, which prompted our redefinition of it in 2013 as the operational DBMS market (see "The OLTP DBMS Market Becomes the Operational DBMS Market"). With the introduction of NoSQL and Hadoop in support of unstructured data in transactions and the viable use of in-memory computing, many organizations are beginning to use these new DBMS engines for specific use cases, such as global scalability of Web applications.

As there are now many new entrants, including small and less mature vendors, the market appears to be split in two, although compared with the situation in 2013, the space between its two groups has reduced. In one group are the innovative new vendors. In the other are the traditional, strong, mature leaders. The reason for most of the vendors being below the midline in the Magic Quadrant is that they support only one or two of the DBMS models (which include key value, graph, relational, table-style and document-style), and only one or two of the multiple data types (such as structured [relational], unstructured, XML, interaction and observation). The Leaders support a wide range of models and data types in a scalable, highly available environment, which is one reason for the considerable space between several of them and the newer vendors.

Another major focus for vendors in this market is support for in-memory computing. Most vendors are beginning to add this functionality to their DBMSs, some with an in-memory-only model. Due to its inherent speed, in-memory computing is becoming necessary for the processing of interaction and observation data integrated into transactions. Most of the traditional vendors have introduced an in-memory DBMS, generally in support of analytics. This will eventually become the preferred model for all DBMSs. The one form of memory not well adapted to data is flash when used as addressable memory and not as a form of fast disk replacement. NAND flash is slower than DRAM, but it is more efficient when used as addressable memory than as a disk block cache.

As the new vendors mature and offer a wider range of functions, the operational DBMS market will become more homogeneous and commoditized (see "IT Market Clock for Database Management Systems, 2014").

This Magic Quadrant deals with the key information management capabilities for transaction processing. It should therefore interest anyone involved in defining, purchasing, building or managing a transaction-processing environment — notably, CIOs, CTOs, infrastructure managers, database and application architects, database administrators and IT purchasing managers.

For this Magic Quadrant, we based our analysis on information gathered from interactions with Gartner clients over the past 12 months and our survey of the vendors' reference customers, performed during July and August 2014.³ We also considered earlier information and any news about vendors' products, customers and finances that arose during the analysis time frame.

The OLTP DBMS market, from which the operational DBMS market evolved, was very mature in the early years of the 21st century. However, as Internet usage and availability grew, so did the applications necessary to support the associated, growing infrastructure. Consequently, over the past five years, many new vendors have entered this market with products to support the specialized applications required by a new and global business arena.

Many drivers of innovation are widely accepted. New forms of data — that were previously difficult or impossible to capture — have become available from connected devices (the Internet of Things), such as smart meter data and machine or device data; we call this "observation data." The pervasive use of personal devices and social media has also become a source of social- and business-related data; we call this "interaction data." These new forms of data must now be used not only for analytics, but also within transactions. Data from vendors' reference customers confirms this change, as 75% of the respondents to our survey use interaction data in transactions, and over 50% use observation data in transaction processing.³

In terms of hardware, new devices, servers, and networking, memory and storage options (to name but a few) have proliferated. These both enable and require new ways to process the data they create or support. For in-memory DBMSs, the amount of memory available on individual servers is already reaching 32TB to 64TB. Organizations need to capture both structured and unstructured data for use in transactions. Furthermore, they must use the data from transactions, observations and interactions in real time for decision processing as part of, not separately from, the transactions. This process is the definition of HTAP (for further details, see "Hype Cycle for In-Memory Computing, 2014").

To support the new operational DBMS market, many new vendors have emerged with innovative DBMS engines that support transactions on a global scale, real-time transactions integrated with analytics, streaming data transactions, and more. These new vendors are single-minded in terms of direction. Once their ideas are proven, more mature vendors will feel obliged to adopt some of this technology.

The new vendors' activities include the use of JSON for data structures in applications; new and less restrictive forms of consistency (allowing for eventual consistency); larger amounts of DRAM for in-memory DBMSs; and new file structures that differ from relational ones, such as those for key-value and document-store file systems. Many of these vendors support these NoSQL systems for simplicity, agile development, support of unstructured data types, scalability and performance. Already, an increasing number of mature DBMS vendors are adopting these technologies in their systems.

Although we exclude cloud-only delivery from this Magic Quadrant, the cloud is being widely adopted as a delivery platform in the operational DBMS market. Over the next few years, we expect most vendors to offer cloud versions of their DBMS products. These will range from simple offerings of support for infrastructure as a service and cloud hosting, to full cloud DBMS platforms with elasticity and multitenant capabilities. As the operational DBMS market matures, cloud deployment — and especially hybrid deployment — will become a criterion of importance as it offers users an additional platform choice.

출처 : http://www.gartner.com/technology/reprints.do?id=1-23A415Q&ct=141020&st=sb