Some observers are questioning the reliability and performance of high-capacity, storage-intensive disk drives including ATA (also known as parallel ATA and PATA) and SATA (Serial ATA), as well as whether they should be used in enterprise environments. But a closer look reveals that this is not entirely true. Used properly, SATA is reliable technology. The issue is not the technology itself, but rather the way it is used -- or more...
Trends regarding capacity-intensive storage include higher capacity disk drives (moving from 400 GB to 500 GB and larger capacity), reduced footprint and faster support (shifting from 5,400 rpm to 7,200 rpm and even 10,000 rpm) for native command queue, which is similar to the SCSI command tag queue for better performance.
Storage array vendors are adding more intelligence into their controllers and firmware to leverage high-capacity disk drives, in order to prolong their life and improve service delivery. Some storage controller and storage system vendors have introduced new data protection techniques to support high-capacity disk drives. These data protection techniques include multi-disk parity schemes using RAID-6 (dual parity). Other examples include predicative copy and data migration schemes to move data from a disk that is likely to fail (avoids or mitigates the overhead of a drive rebuild).
Vendors such as Copan Systems refer to this as disk aerobics, where they power down disk drives that are not being used to prolong the drive life and stay within disk drive duty cycle parameters. EMC Corp., Engenio Information Technologies, Hitachi Data Systems, HP, IBM, Network Appliance Inc., Sun Microsystems Inc., StorageTek, Xiotech Corp. and Xyratex are some of the vendors leveraging good packaging including power, cooling, noise and vibration dampening to help improve the performance and reliability of high-capacity disk drives. One way to maximize the performance and availability of a high-capacity disk drive is to use it for what it was designed to do. Most challenges and issues related to capacity-intensive disk drives (at least the issues I hear about) are tied to product configuration and use. In other words, the technology is being made to perform tasks it was not intended to be used for.
Here are some tips and recommendations:
- Align the applicable disk drive technology to the specific application need, requirements and service characteristic requirements. For example, use capacity-intensive disk drives for storage-intensive applications and high-performance disk drives for I/O intensive applications.
- Select capacity-intensive storage (ATA, SATA and FATA) with packaging that includes airflow, power, vibration isolation and redundant, hot-swappable components to maximize availability and performance.
- Ask your vendor what their configuration rules are and if they have any guidelines, best practices, or support material to help you get the maximum value out of the technology.
- Look for storage arrays that are ATA-, SATA- or FATA-aware and that maximize duty lifecycle of these drives.
- Check with your vendor as to what type of disk drive they are using -- be it a desktop or an enterprise-class drive -- and the qualification and selection process of which drives they are using. Note: Not all ATA and SATA disk drives are the same; there are different classes and categories with the differences in some disks being how they are tested.
- Talk with your vendor to see if disk drive failures are real (drive actually failed, data is lost) or artificial (data is intact, disk did not actually fail). Many drive failures are actually artificial in that a storage array senses that something might be wrong, so it flags the disk drive as 'bad.' While there could be something wrong with the disk drive itself, it is also possible that the sensitivity of the storage array might be such that it is proactively taking a safe approach, i.e., not taking any risks with regard to data integrity.
- Avoid making your capacity-intensive disk drives perform unnatural acts or using them for things that they were not designed for, namely I/O and performance-intensive applications with continuous I/O activity. If you have I/O or performance-intensive applications, then point those to Fibre Channel (FC), SCSI, or SAS-based disk drives.
- You can expect to get what you pay for, so pay a bit more for a premium SATA disk drive and associated packaging. Look beyond the initial dollar per GB and look at how long you plan on using the storage; in other words, look at the total cost of ownership.
- Extended length disk-drive rebuilds are not unique to ATA and SATA disk drives. However, they are more in the spotlight given their relative performance compared to FC, Parallel SCSI and SAS disk drives as well as their larger capacity. FC disk drives are available with 300 GB, and larger capacity drives are anticipated.
- Consider storage arrays that support enhanced parity schemes and accelerated disk-drive rebuild and predictive disk-drive copy features to minimize exposure during extended length disk-drive rebuilds.
- Pay attention to how many I/O streams are accessing a SATA disk drive at the same time, as well as what type of I/O operations (reads, writes, random or sequential) along with short burst I/O or sustained, large data transfers.
There are many myths and a lot of fear, uncertainty and doubt about SATA and high-capacity disk drives in the industry today. The key is to become more aware of its strengths and weaknesses, and how to use it as part of a tiered storage environment. You can learn more about what type of disk drive to use for different applications in the free white paper "What Type of Disk Drive to Use", which is available from The Evaluator Group.
For more information:
About the Author: Greg Schulz is a senior analyst with the Evaluator Group responsible for distributed storage, networking, and associated platform management software. Greg is also author of the book "Resilient Storage Networks" (Elsevier). Greg can be reached at email@example.com
This article originally appeared on SearchStorage.com.