TL;DR: At one time, organizations planning their cloud strategy adopted an either-or approach: Either store and manage data on a secure private cloud, or opt for the database-as-a-service model of the public cloud. Now companies are realizing the benefits of both options by adopting a multicloud strategy that places individual applications on the platform that best suits them.
In IT’s never-ending quest to improve database performance and reduce costs, a new tactic has surfaced: multicloud. Rather that process all database queries on either the private cloud or public cloud, shift the processing to the platform best able to handle it in terms of speed and efficiency.
InfoWorld’s David Linthicum explains in an August 5, 2014, article that a multicloud architecture “gives those who manage large distributed databases the power to use only the providers who offer the best and most cost-effective service — or the providers who are best suited to their database-processing needs.”
Managing the resulting complexity isn’t as daunting as it may sound, according to Linthicum. In fact, a cloud-management system could soon become a requirement for IT departments of all sizes. Product lifecycle management (PLM) expert Oleg Shilovitsky claims in an August 5, 2014, article on BeyondPLM.com that three trends are converging to make distributed database architectures mandatory.
The first trend is the tsunami of data that is overwhelming information systems and pushing traditional database architectures to their physical limits. The second trend is the increasingly distributed nature of organizations, which are adopting a design-anywhere, build-anywhere philosophy. The third trend is the demand among users for ever-faster performance on many different platforms to keep pace with the changes in the marketplace.
Multicloud: More than simply pairing public and private
In a July 12, 2013, article, InfoWorld’s Linthicum compared the process of adopting a multicloud strategy to the transition a decade or more ago to distributed internal systems customized to the specific demands of the business. A key to managing the increased complexity of multicloud systems is carefully choosing your service provider to ensure a good fit between their offerings and your company’s needs.
Three key considerations in this regard are security, accessibility, and scalability. These are three areas where the Morpheus database-as-a-service shines. In addition to lightning-fast SSD-based infrastructure that increases IOPs by 100 times, Morpheus provides real-time monitoring for identifying and optimizing database queries that are impeding database performance.
Morpheus offers ultra-low latency to leading Internet peering points and cloud hosts. Additionally, fault tolerance, disaster recovery, and automated backups make Morpheus a unique Database as a service. You connect to your databases via secure VPC. Visit the Morpheus site for pricing information or to create a free account during the beta period.
Mixing old and new while maximizing adaptability
Businesses of all types and sizes are emphasizing the ability to shift gears quickly in anticipation of industry trends. No longer can you simply react to market changes: You must be there ahead of the competition.
A principal benefit of the multicloud database architecture is flexibility. In an August 25, 2014, article on Forbes.com, IBM’s Jeff Borek highlights the ability of multicloud databases to leverage existing IT infrastructure while realizing the agility, speed, and cost savings of cloud services.
A typical multicloud approach is use of the private cloud as a point-of-control interface to public cloud services. MSPMentor’s Michael Brown describes such an architecture in an August 27, 2014, article.
Many companies use a private cloud to ensure regulatory compliance for storing health, financial, and other sensitive data. In such systems, the private cloud may serve as the gateway to the public cloud in a two-tier structure. In addition to providing a single interface for users, the two levels allow applications and processes to be customized for best fit while keeping sensitive data secure.
A multicloud-application prototype: Managing multiple application servers
There’s no denying that managing a distributed database system is more complicated than maintaining the standard top-down RDBMS of yesteryear. In a July 23, 2013, article on GitHub, German Ramos Garcia presents a prototype multicloud application development model based on the Hydra service. The model addresses much of the complexity entailed in managing multiple application servers.
The web application is first divided into static elements (images, Javascript, static HTML, etc.), dynamic elements on a backend server, and a database to support the backend servers.
A prototype multicloud application architecture separates static, dynamic, and database-support servers.
The distributed architecture must provide mechanisms for controlling the various servers, balancing traffic between servers, and recovering from failures. It must also control sessions between servers and determine where to store application data.
An alternative approach to multicloud management is presented by Mauricio J. Rojas in a blog post from March 25, 2014. The model Rojas proposes is a mash-up of management tools from many different cloud services.
Management tools for distributed cloud-based databases should focus on user needs and offer best of breed from various providers.
Rojas recommends creating a single set of management components for both the public and private clouds. This allows you to “create the same conditions in both worlds” and move seamlessly between the public and private domains.
In addition to security, important considerations in developing a multicloud management system are auto-scaling and high availability. With the Morpheus database-as-a-service, you’re covered in all three areas right out of the box–even Pinal Dave, the SQL Authority uses Morpheus. Make Morpheus a key element of your multicloud strategy.
