lundi 11 novembre 2013

Big Data as a Service: Business customers need more than analytics

A lire sur:

Gina Narcisi Published: 07 Nov 2013

Analytics and search and discovery capabilities have become table stakes for cloud providers offering big data services. To differentiate in the emerging Big Data as a Service market, providers must offer added services like secure, high-bandwidth network connectivity for customers accessing and processing their large data sets, as well as big data consultancy services for planning and implementation assistance.
"Customers need more than just a place to store their big data," said George Gilbert, co-founder and partner of management consulting and research firm TechAlpha. "Providers need to have the fiber in place to move data, and … at this stage, it's very important to offer consulting services on top, because big data management software isn't something that many business customers have experience with."

Savvis Big Data as a Service: Hardware, software and consulting for business customers

Savvis Inc., a managed hosting, colocation and Infrastructure as a Service (IaaS) provider, recently announced Savvis Big Data Solutions, a suite of managed, dedicated hardware and software services that offer data storage, integration and retrieval capabilities through Savvis' IaaS platform, said Milan Vaclavik, senior director and solution lead for Savvis' Big Data Solutions.
Savvis Big Data Solutions offers scalable compute power over a secure network connection owned by Savvis, Vaclavik said. Savvis' initial Big Data Solutions release is a single-tenant cloud offering with hardware dedicated to individual customers. "Savvis isn't offering [Big Data as a Service (BDaaS)] in a public cloud environment at this point because of data privacy and security concerns for some customers -- like financial services -- who [Savvis] is aggressively targeting right now," he said. 
The big data services also offer software licensing and operations management for tools that many enterprises are already using for storage -- Cloudera and MapR platforms -- which are based on Apache Hadoop, an open source, programming framework for processing large data sets.

More on Big Data as a Service:

Big data analytics delivered via SaaS
Making big data a smaller problem with cloud object storage
Cloud analytics offer relief for big data
Savvis Big Data Solutions also includes planning, implementation and consulting services that will provide customers with environment design, security planning and project management, Savvis said.
"Depending on where a customer is with their use of Hadoop, they may take advantage of the consultancy services," Vaclavik said. "For example, they may not want to manage [Hadoop] in their own data centers anymore, in which case, Savvis can help with migration services."
Big data analytics are not included in the BDaaS portfolio yet, but will be in the future, he said. "We will be partnering with analytics providers and systems integrators to make big data analytics part of the solution for our customers who aren't looking to do [analytics] in-house."
For public cloud IaaS providers who are not trying to compete head-on with Amazon -- such as Savvis, IBM and Verizon Terremark -- big data services are a good way to differentiate themselves within the market, said Dave Bartoletti, senior analyst at Cambridge, Mass.-based Forrester Research Inc. "These providers can't just throw up empty infrastructure like the big providers have. They should develop infrastructure that helps customers get to what they want to do with their infrastructure faster with business intelligence tools," he said.

Barriers to the delivery of big data cloud services

As a subsidiary of telecommunications provider CenturyLink, Savvis can also differentiate its big data services with network connectivity. Not every cloud provider will have the pipes necessary to receive and move big data sets, TechAlpha's Gilbert said.
Some providers -- like Amazon -- have historically resorted to shipping large hard drives out to customers to upload their data and ship it back, as shipping data via ground delivery was faster than uploading large data sets over the Internet, Gilbert said. "The network is typically the big constraint, and moving data is a big expense," he said.
In addition to bandwidth constraints, tying various big data services -- like search and analytics -- into billing and management software on the back end is a fixed, upfront expense that will require the same amount of effort and cost for any provider, he said. "Larger providers will have more customers to sell the services to and spread their investment over, making it inherently more profitable."

Aucun commentaire:

Publier un commentaire