How Software Defined Storage Can Improve Your Entire Company
by Andrew Mullen on February 1, 2017
Software Defined Storage (SDS) is having a substantial and growing impact in today's corporate data centers. A survey conducted by ESG found that 60 percent of companies are already committed to employing SDS as an integral part of their storage infrastructure, while another 23 percent are showing a strong interest in doing so.
The reason for that enthusiasm is the benefits this new approach to storage infrastructure can provide to IT operations struggling to handle an exponentially growing outpouring of data, while budgets remain nearly stagnant. SDS provides advantages over traditional storage solutions in areas such as flexibility, scalability, and reduced storage management complexity, as well as significant reductions in hardware costs. So, when it comes to seriously considering a move toward implementing SDS in their data centers, many IT managers are already on board.
But the benefits of a software-driven approach to storage are not confined to the data center. SDS can have a positive impact on the entire organization. In this article we'll take a look at some of the ways that can happen.
What, Exactly, Is SDS?
One of the obstacles that has slowed adoption of SDS in some organizations is fuzziness about exactly what the term means. As is often the case when new technologies begin to gain momentum in the marketplace, some vendors have attempted to repurpose existing products by extracting the software component and labeling it as SDS. That's led to some confusion about just what "software defined" really signifies.
The Storage Networking Industry Association (SNIA) defines SDS as "virtualized storage with a service management interface." What that means in practice is that with SDS, storage is accessed and managed by software rather than by direct interaction with the underlying hardware devices. Users are presented with a uniform, "single pane of glass" interface regardless of the nature, configuration, or geographical locations of the actual storage resources.
SDS software manages all the devices it controls as a single pool of storage. One important advantage of this approach is that the intelligence that implements the sophisticated functionality on which a modern storage system depends, features such as data deduplication, replication, snapshotting, backups, and disaster recovery, resides in the software rather than at the hardware level. This means that inexpensive COTS x86 servers and commodity drives can be used in place of the specially designed, and far more costly, dedicated storage hardware employed in traditional storage solutions. As importantly, this approach allows an organization's entire storage infrastructure to be managed in a coherent fashion from a single point of control.
SDS Helps Reduce Both CapEx and OpEx Costs
Traditional dedicated storage arrays are expensive to purchase and to support. SDS makes it possible to forego use of these proprietary devices in favor of less expensive commodity hardware. The result is that in a SDS data center, CapEx expenditures for acquisition of storage arrays, as well as OpEx spending for maintenance and support, can be substantially reduced.
In addition, because SDS can automatically provision resources as needed through software, rather than by IT staff having to procure and physically attach new hardware, storage resource utilization is greatly improved. One study estimates that 10 million servers are sitting idle in corporate data centers, tying up $30 billion in useless capital expenditures. The reason is that in the traditional storage model, managers must purchase excess capacity to be ready to accommodate unexpected surges in demand. In many cases, with technological innovations appearing at an ever-increasing rate, those standby devices become obsolete before they are ever used.
But with SDS, the software can be programmed to automatically assign resources from anywhere in a potentially world-wide storage pool as needed, including those provided through a private or public cloud, without the necessity of new CapEx expenditures. In addition, since SDS can cheerfully make efficient use of legacy hardware that an organization may already have on hand, the effective service life of existing capital assets can be extended. Because storage resources are being used more efficiently, the organization's cost per TB of storage can be dramatically lowered.
The ability to use a mix of storage technologies, some inexpensive but relatively slow and "dumb," and some that deliver high performance but at a high price, is one of the most attractive features of SDS. By properly choosing devices and media to match the organization's workloads, sizable reductions in total cost of ownership can be achieved. In fact, Gartner reports that "SDS solutions can reduce TCO by 50% or more without sacrificing performance, data services software robustness or availability service-level objectives."
SDS Can Provide Decision Makers With Better Information
Information silos occur when the data repositories used by various departments or business units within an organization are effectively segregated from one another. For example, a company's ERP (enterprise resource planning) system may run off a separate database, housed in a separate storage system, from those used by its CRM (customer relationship management), sales, or human resources applications. This inability to share related types of information can create data integrity issues, and makes using that data to obtain organization-wide insights extremely cumbersome and error-prone.
Another example of the pernicious effect of information silos occurs when branch locations each maintain their own IT facilities on site, generating, updating, storing, and backing up data locally. Making ROBO (remote office/branch office) information available to decision makers elsewhere in the organization can be a severe challenge. All too often, incompatible copies of what are assumed to be identical datasets or files exist in different locations within the company. Updates made at one location are not distributed to the rest of the organization in a timely fashion, so that managers find themselves, often unknowingly, making decisions based on outdated, incorrect, or inconsistent information.
With SDS, a company's entire collection of data, no matter where it is generated or used, can be consolidated into a single resource controlled and accessed through the SDS interface. This allows use of analysis tools that span the organization's entire storage infrastructure, making that information immediately available to decision makers.
No local copies of data are maintained, eliminating the requirement for backup and disaster recovery capabilities in remote locations. Users in ROBOs interact directly with the data held in the organization's central data repository so the consistency and freshness of the information is automatically maintained. The SDS software can be designed to use sophisticated local caching and intelligent file-locking technologies to insure that changes are immediately propagated throughout the organization, and that data corruption due to different users attempting to make simultaneous updates to the same files cannot occur.
SDS Increases Business Agility
In today's business environment, conditions can change with lightning speed. Keeping up with changes, and taking advantage of opportunities as they occur is often a function of the enterprise's ability to derive actionable business intelligence from an ever-increasing torrent of data. That means the organization's storage infrastructure must be able to quickly adjust to changing capacity demands in order to reliably capture, retain, and deliver for analysis the unprecedented amounts of data being generated every day.
With traditional storage solutions, responding quickly to new workloads and storage requirements can be a challenge. Because expanding capacity normally involves capital expenditures for new hardware, unanticipated surges in the demand for storage may be difficult to quickly accommodate. And once new hardware is purchased, configuring each unit and incorporating it into the storage network must be done manually, a time- and cost-intensive process.
SDS allows companies to quickly ramp up their data storage infrastructure to meet suddenly exploding demands for storing and analyzing information. With SDS, provisioning to meet expanded demand can be an automated software function that takes minutes rather than days or weeks. This allows businesses to be much more agile in acquiring and analyzing the data they need in order to be responsive to changing conditions.
SDS Helps Keep Mission-Critical Information Secure and Safe
One area in which SDS can make a major contribution to the health of the enterprise as a whole is in data security. With SDS the organization's entire storage infrastructure is managed through a single interface, and with one consistent set of policies. This allows the company's IT staff to apply security best practices to the system as a whole, rather than having to do so for each individual storage array or subsystem. Best-in-breed implementations of functions such as data encryption, backup, replication, and disaster recovery can be applied across the board.
Access permission policies can be administered for the entire storage system through the SDS interface, increasing consistency in their application. Storage administrators can gain insight into the operation of each element of the system that would be very difficult to obtain with traditional solutions in which each storage subsystem must be addressed as a discrete unit. Monitoring and analysis can be done through the SDS console to highlight attempts to penetrate any part of the storage system.
The result of this ability to apply a consistent set of policies to the storage system as a whole is that data security can be significantly enhanced with SDS.
SDS also excels at protecting against data loss or corruption due to equipment failures. Because SDS is designed to employ COTS servers and drives, unit failures are expected and planned for. Data is automatically replicated in several locations, and if a storage array goes offline, the system can automatically reconstitute itself without disrupting operations. When failing units are replaced, they are automatically incorporated into the system, all without IT staff having to manually direct the restoration of data. After implementing SDS at a campus of the University of Hawaii, systems programmer Paul Ryan confirmed that point. "The layer of abstraction that SDS provides is important in that we don't have to manually recover drives," he says. "The cluster does that."
SDS Greatly Reduces Storage Management Complexity
With SDS storage administrators can manage the organization's entire storage infrastructure from a single location through a single software interface. Although actual storage units may be located in different parts of the world, may have different hardware designs and configurations, and may even be of different media (hard disks vs flash drives, for example), an administrator can deal with each element of the system in a consistent way.
The SDS is able to automate many functions according to policies that are keyed to specific applications and workloads, relieving the IT staff of time-consuming operational responsibilities.
The simplification of storage management that is inherent in SDS provides company-wide benefits by relieving IT staff of mundane maintenance and operational tasks. The policy-driven automation of storage system functions that SDS implements is actually more reliable and less error-prone for many of the routine but complex tasks IT staff were formerly required to perform on a daily basis. The result is fewer disruptions to mission-critical applications and workloads, and greater satisfaction by both internal users and customers.
These are just a few of the ways SDS can benefit an entire company. If you'd like to know more, please visit our Next Generation Software-defined Storage page.