twobee - Fotolia
The database-in-the-cloud market, poised for stellar growth, has channel partners vying for a piece of what's expected to become a $1 billion-plus opportunity.
Database as a service (DBaaS) has been around for years but was limited mainly to test and development environments. With expanded uses for this cloud service, however, the market has taken on new life. For example, DBaaS has become a key launch vehicle for emerging, non-relational database technologies such as NoSQL and other big data-oriented software. But DBaaS also sees action among relational systems such as MySQL. Deployments for both types of databases may occur in public or private clouds.
The channel ecosystem forming around DBaaS has found a number of roles for it. While hosting a database in the cloud has functioned as the primary service line, companies now offer additional services around the core hosting offering, including database administration, performance diagnostics and optimization.
The integrators, managed services providers and cloud specialists that enter this market could be in for quite a ride. 451 Research LLC, a market research firm based in New York, estimates that DBaaS providers will generate $1.8 billion in revenue by 2016, projecting a compound annual growth rate of 86%. MarketsandMarkets offers a similarly optimistic view. The company pegs the cloud database and DBaaS market at $14.05 billion by 2019, forecasting a 67.3% annual growth rate.
Chip Childersvice president of product strategy, CumuLogic Inc.
The main appeal of DBaaS is much the same as it is for other "as a service" offerings: It lets organizations deploy the technology faster, and it frees them from the drudgery of ongoing maintenance and support.
"There's a huge opportunity for database teams to get themselves out of the day-to-day, boring work of deploying and managing these databases," said Chip Childers, vice president of product strategy at CumuLogic Inc., a Santa Clara, California, company that provides a DBaaS software platform.
NoSQL's cloud boost
A number of channel companies aim to take advantage of the growing interest in DBaaS, and much of the recent activity surrounds nontraditional databases.
INetU Inc., a managed cloud and hosting company based in Allentown, Pennsylvania, in June expanded its database line of business to include Aerospike, an open source, in-memory NoSQL database. The company also hosts and manages Oracle, Microsoft SQL Server and MySQL databases.
David Fowler,vice president of marketing at INetU, refers to his company's offering as "cloud database services" as opposed to "DBaaS." Fowler said he views DBaaS as a hosted service in which the database service is already set up and the customer connects to configure their database environment. With INetU's cloud database services, including the Aerospike hosting solution, the company spins up a database to work with a customer's applications in a public- or private-cloud setting. The hosting environment, he said, is tuned for the customer's particular applications rather than set up to operate independently of those applications.
"We're not just waiting for a user to connect," he said. "We spin up Aerospike as part of your solution."
Hosting service provider Rackspace, meanwhile, has been adding NoSQL and big data cloud services, in some cases with the help of recently acquired companies. The San Antonio, Texas, company in 2013 purchased ObjectRocket, which now provides the basis for Rackspace's hosting solution for MongoDB, a widely adopted NoSQL database. Another 2013 acquisition, Exceptional Cloud Services, brought with it a cloud product for managing instances of the Redis NoSQL database. Rackspace plans to launch a Redis-in-the-cloud offering on the ObjectRocket platform. That service is scheduled to go live July 31.
Rackspace's lineup also includes a Hadoop as a service offering. Hadoop provides software for the distributed processing of large datasets on a cluster of servers. Customers may use the Apache HBase database in a Hadoop deployment, but more often they would export the results of a Hadoop job back into a relational database. The company supports MySQL, SQL and Oracle on the relational database side.
The speed and flexibility of cloud hosting, however, suits the fast pace of innovation in the non-relational database field.
"It allows us ... to react to the needs of those types of platforms," said Sean Anderson, manager of data services at Rackspace.
And in another NoSQL move, CumuLogic in June launched the MongoDB Edition of its DBaaS software platform. The company said the software lets IT organizations create an on-premises MongoDB as a service offering for application developers as well as quality assurance/testing and operations groups. CumuLogic sells its DBaaS software to enterprises and service providers.
"We run into companies that have a lot of questions [about] the right way to deploy MongoDB," Childers noted. "Really, one of the hurdles to adopting a database technology that is perhaps newer is the knowledge gap."
In the case of MongoDB, an enterprise's developers may begin working with the database before the broader IT department is ready to deal with it. Database administrators, for example, will typically need to get up to speed on learning the steps of provisioning a MongoDB cluster. A prebuilt DBaaS platform, however, eliminates much of the heavy lifting, Childers suggested.
"[MongoDB brings] us into deals where the enterprise needs to find a way to make it easier … to operationalize MongoDB technology that developers have already adopted," he explained.
Customers' lack of familiarity with unconventional databases creates an opening for channel partners to extend their cloud database services beyond hosting. But the newer database technology isn't the only factor driving the wider set of services. Some enterprise IT managers and applications developers are simply looking to offload database chores, regardless of the technology's vintage, according to industry executives.
Derek Shoettle, CEO of Cloudant, looks at the expansion of services as part of the market's evolution. Cloudant was an early DBaaS provider, having launched in 2009. The company, which seeks to manage the data layer for mobile and Web application developers, was acquired by IBM in March.
Schoettle said earlier hosted services compelled developers to make decisions about the database, the underlying architecture and the storage format. Cloudant's NoSQL DBaaS, however, aims to counter this do-it-yourself approach to databases. Cloudant provides database hosting but also takes on such chores as load balancing, scaling the size of clusters and creating a high-availability environment.
"Developers don't want to make all of the decisions and assume responsibility for picking the underlying data repository," Schoettle said. "They just want to worry about the API [application program interface] and performance and service quality so they can build the apps."
Rackspace's Anderson described a sliding scale of offerings, ranging from DIY databases to "not just database as a service, but data as a service."
As for the latter offering, the idea is to cover database administration, cluster management and patching and updating. Anderson said ObjectRocket provided a platform that automated many of the pain points around database administration. That model will apply to the entirety of Rackspace's database hosting operations going forward.
"We are in the process of applying the ObjectRocket method to all of our platforms," Anderson said.
Customers will be able to upload their data and not worry about the management duties around the database engine, Anderson added.
Fowler said INetU's approach with its Aerospike cloud database service is to create a tuned environment that lets an application take full advantage of a NoSQL database's high performance. Once a customer links its application to the Aerospike service, INetU conducts an analysis using synthetic transactions to make sure the performance is actually there, Fowler said. Synthetic transactions, which simulate customer activity, are used to determine how a system is performing under various realistic loads.
INetU also uses this database-tuning approach in its SQL business. Fowler said many companies need help with database implementation. Application performance issues, in many cases, can be traced to the database layer, he noted. He said INetU diagnoses problems via synthetic transactions and then performs the necessary tuning.
The database and its surrounding cloud services occupy a key junction between the application, the underlying infrastructure and a customer's business objectives.
"The database sits right in the center," Fowler said. "I see that as the critical component."
How to plan your DBMS strategy