Big savings with MAID 2.0 storage technology
Network World - For years, companies have been deploying massive arrays of idle disks (MAID) technologies to reduce reduces energy consumption by putting power-hungry disk drives to sleep when they're not being used.
Early versions of MAID products -- whether you think of them as first-generation MAID, MAID 1.0 or "old MAID" -- have two shortcomings, however. First, spun-down drives typically take several minutes to spin back up. That's an unacceptable length of time and results in an unacceptable loss of performance for most applications. Second, most MAID 1.0 providers support only a limited percentage of active disks at any given time, frequently as few as 25%.
MAID 2.0 broadens energy-saving options by adding multiple modes -- instead of the binary on/off approach of MAID 1.0 -- that are designed to accommodate different types of data. For instance, files such as financial records and medical test results need to be stored for several years or even longer, but their data doesn't change. An X-ray, for example, needs to be stored, but it typically is not accessed frequently if at all, and it is never altered. Such fixed-content data is ideally suited for MAID 2.0 applications.
Disk drives consume 80% of the power used in data storage, and unless technologies focus on solving this specific issue, they're missing the point. Both MAID 1.0 and MAID 2.0 meet this critical criterion, but MAID 2.0 delivers the additional benefit of performance assurance. The MAID 1.0 approach of turning off spinning disk drives can achieve significant power savings, but because this practice has such a dramatic and negative effect on application performance, it is not practical in most circumstances. The unwavering goal of MAID 2.0 technologies is to deliver energy savings to the broadest set of applications without affecting their performance.
MAID 2.0 products achieve this goal by operating at user-defined and application-dependent energy-saving levels. A Level 1 setting signifies the fastest response time that still makes energy savings possible. A Level 1 setting would be ideal for busy organizations such as hospitals, which need fast access to patient X-rays or records but can save energy when there are idle periods.
Many people incorrectly believe that round-the-clock operations' data centers never experience idle time. However, according to a recent information-access study by researchers at the University of California, Santa Cruz, 90% of all stored data was not touched during a three-month time frame. Being open for business doesn't mean that all of an organization's storage disks need to be spinning at full speed.
A typical user might elect to put his storage devices into Level 1 MAID 2.0 after 15 minutes of inactivity. At that point, the read/write heads will be unloaded automatically. This simple step, transparent to the user, reduces air resistance and can provide an energy savings of approximately 20%. It maintains a subsecond response time without a power spike for the first I/O request. All subsequent I/O requests are handled instantaneously.
- 10 Hot Big Data Startups to Watch
- 11 Unique Uses for Google Glass, Demonstrated by Celebs
- How to Export Your Google Reader Account
- How to Better Engage Millennials (and Why They Aren't Really so Different)
- Telltale signs of ATM skimming
- 20 security and privacy apps for Androids and iPhones
- Big screen con artists: 7 great movies about social engineering
- IT Certification Study Tips
- Register for this Computerworld Insider Study Tip guide and gain access to hundreds of premium content articles, cheat sheets, product reviews and more.
- How Apollo Group Evaluated MongoDB Apollo Group, best known as the parent company of the University of Phoenix, sought to build a cloud-based learning management platform and needed...
- Total Cost of Ownership Comparison of MongoDB & Oracle In this white paper, we compare the total cost of ownership (TCO) of MongoDB and Oracle. It can be faster and cheaper to...
- Big Data: Guidelines for the Enterprise Decision Maker In this paper, we begin with a description of Big Data and the data management landscape. Next, we describe examples of customers innovating...
- Storage Infrastructure as a Service: The Best of Cloud & On-premises Storage Read this white paper and discover the benefits of moving from traditional storage to Storage Infrastructure as a Service.
- Gartner Key Data Protection Challenges Analyst Video Shifting market dynamics, new delivery models and environments, data created at the endpoints, and flatling budgets mean the data center is undergoing a...
- Becoming An Analytics Driven Organization Join us on Tuesday, June 18, 2013, 11:00 AM EDT and learn how your agency can create an analytics culture that will enable... All Data Storage White Papers | Webcasts