Home
StorCase Technology Safest Place for Your Data
Company/Contact Info Support Resources



Data Integrity - a Storage Priority

More devastating than an occasional checksum error or an undetected virus are the potential threats to today's hard disk drives. With the continual increase in demand for system speed and storage capacity, complete drive failures resulting during even the most fundamental operations are now a common reality.

To overcome the growing "failed drive syndrome" experienced by storage systems, expensive compensation technologies such as RAID or mirroring are often employed. These "after-failure" methods not only incur the costs of replacing the failed drives, but they also assume and require the attention of already-taxed network administrators. It becomes apparent that the concept of drive protection and thus failure prevention should be of top priority when selecting today's storage solutions.

For a more specific understanding of why drive failures have become more common, and how some enclosures are more likely to prevent these failures, it is helpful to consider the various contributing factors. Why does a drive fail and what recent changes in the industry have made the drive's environment such a critical and growing consideration?

The Growth of a Technology
When IBM first introduced the first hard disk drive in May 1955, it offered unprecedented random-access storage. The product heralded leaps in mass-storage technology and the end of sequential storage on punched cards and paper or Mylar tape. However, the RAMAC (random-access method of accounting and control) disk drive occupied the space of two refrigerators and weighed in excess of a thousand pounds.

Newer technologies gradually allowed the creation of smaller drives, capable of storing previously unimagined quantities of data and performing data transfers at incredible speeds. These accomplishments revolutionized the storage industry such that it is now frightening to consider the proportion of the world's data that stored on hard drives -- much of this data which is maintained with little or no back-up provisions.

Hard drive operation is an electro-mechanical process with rapid moving parts. Manufacturers market disk drive products with accompanying warranties (typically 1 or 3 years) and MTBF (Mean Time Between Failure) ratings. MTBF, a key measure of the reliability of a hardware product or component, is usually listed at a value in excess of 300,000 hours - which appears to be very sufficient. However, the MTBF value assumes that the specified environmental condition requirements for the device are strictly adhered to. For example, specifications may state, "under no circumstance should the disk drive exceed XYZ temperature". If this temperature is exceeded, however, the MTBF can be expected to decrease dramatically. For this reason, the value of maintaining favorable environmental conditions, and therefore a meaningful hardware MTBF, becomes very important. There are several relevant storage enclosure design aspects that should be considered.

Enclosure Design Considerations

Heat Dissipation
The demand for greater storage capacities and faster transfer speeds, accompanied by the natural consumer demand for smaller devices, has clearly led us to today's smaller and more powerful storage products. Today, spindle speeds of up to 15,000 revolutions per minute are not uncommon. This represents a threefold increase in just five years. These high-performance disk drives, when placed in a highly data-intensive application, can substantially impact the amount of heat generated by the storage enclosure. Especially in mission-critical environments, if this heat is not dissipated effectively, the resulting data errors or premature device failure can be devastating.

A properly designed storage enclosure will employ construction materials such as steel, which provides ideal thermal properties for quick heat dissipation, and will also provide ample space and robust cooling mechanisms to allow continuous heat dissipation. Lower cost, mass-produced "fan"s are typically no longer adequate; draft induced, turbine blowers are a better choice to ensure that today's hard drives experience continuous and adequate cooling. In addition, high-pressure blowers should incorporate safety features such as auto-speed adjustment and failure reporting, as well as the capability to function in a dual redundant, hot swappable system.

Power Management
Another critical storage environment component is the system's power supplies and the management circuitry supporting them. Disk drives must receive continuous, smooth and sufficient power to operate properly. However, the amount of power required by a drive at any one time varies depending on the function it is performing. A well-designed power system not only provides a high enough wattage, but it also assures that appropriate levels of power are applied to the devices at all times.

Power supplies should be hot swappable and should incorporate built-in intelligence to share and balance the "load", ultimately extending the life span of each power supply and preventing wasted power. A well-designed enclosure power system should include built-in provisions that protect the drives themselves from too much voltage or current that can potentially cause immediate or latent drive damage. Further power management design considerations include the ability to isolate the main power supply and drive power inputs, thus protecting the drives from voltage fluctuations elsewhere in the system, or the ability to assure that drives are not powered during hot plugging. All of these design features must be taken into consideration to properly minimize the occurrences of drive errors, damage and/or failure.

Signal Quality
Data integrity of the storage system clearly relies on the continuous quality of the signals transferred between the disk drives and the other system components. With the increased data transfer speeds experienced by today's storage systems, it is particularly critical that storage enclosures employ appropriate system "corrective" measures, such as signal repeaters and re-timers capable of restoring skewed or weak signals prior to passing them along. Experienced PCB (printed circuit board) design is also a necessary consideration to avoid signal "cross-talk" as transfer speeds increase. Signal degradation or interference can lead to reduced performance and data errors.

Rotational Vibration
One can imagine a high-performance household washing machine spinning away at 1,000 revolutions per minute, and how it's not uncommon to feel the whole room shudder. Similarly, a network's individual disk drives are humming away at 15 times that speed, and they also are typically stacked together within a small enclosure space. It follows that 15,000 revolutions per minute can have a serious impact on the disk's ability to maintain consistent operation.

With increasingly high-speed drives dominating the storage market, protecting the drive from rotational vibration (RV) is a key factor in preventing drive damage and failures. Housing hard drives in a rugged enclosure with excellent structural rigidity and drive fit is critical to maintaining the delicate environmental balance necessary to ensure a maximum MTBF. By incorporating preventative RV testing into an enclosure's mechanical design process, it can be assured that the RV guidelines specifically put forth by major device manufactures have been met.

Environmental Monitoring
Finally, organizations concerned with system data integrity should incorporate plans for achieving complete access to the ongoing status of the system's disk drives, the drives' environmental conditions and the status of key enclosure components (such as the power supplies, blowers or RAID controllers.). With options such as on-board and remote monitoring available to measure and provide enclosure status, as well as the status of the entire disk array, there is no logical reason to not be in control of the disk environment at all times.

The Safest Place for Data
Despite being a highly reliable means of data storage, hard disk drives are still mechanical devices with an inherent failure factor that must be anticipated. Unless disk drives are housed within a well-designed environment that is intended to accommodate the stringent demands of the latest drive technologies and applications, the risk of data loss will continue to be significant.

Armed with the knowledge of the key factors necessary for maintaining system data integrity, organizations can make appropriate decisions and then rest assured that their data will be stored and will operate within the safest environment possible. A data centre with well-planned data storage provisions will ensure maximum system up-time and the highest possible levels of data integrity and security, without complete reliance on expensive compensation technologies like RAID. Although disk drives may still fail internally, regardless of the environment in which they have been placed, the overall risk of data loss arising from clearly preventable circumstances can be greatly minimized. This realization should be the very first step taken in the quest to protect precious data.

Copyright © 2006 StorCase Technology, Incorporated. All rights reserved.