Articles: Storage
 

Bookmark and Share

(4) 
Pages: [ 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 ]

File Server Pattern

The next IOMeter pattern simulates a typical file-server load. The request queue depth changes from 1 to 256. Let's first view the results at a queue depth up to 32 requests.

The Barracuda LP ST31000520AS behaves unlike the other HDDs, reaching its top speed at a queue depth of 4 requests. This optimization makes it somewhat faster under real-life conditions but prevents it from showing a higher performance at longer queue depths.

It is the RE4 WD1003FBYX and Caviar Black WD1002FAEX that have the highest results again. They are followed by the Caviar Blue WD10EALX and Barracuda 7200.12 ST31000524AS whose graphs are close to coinciding except that the Seagate is somewhat slower at a queue depth of 1 and 2 requests. The Barracuda 7200.12 ST31000528AS and WD10EALS are even slower, and the Barracuda LP ST1000DL002 fails the test completely.

We’ve got a somewhat different picture when the queue is as long as 256 requests. Seagate's energy-efficient products exchange their places, the Barracuda LP ST1000DL002 now being indifferent to the queue depth (just like the Barracuda 7200.12 ST31000524AS, though). The Barracuda LP ST31000520AS speeds up suddenly as the queue increases from 32 to 64 requests. The RE4 WD1003FBYX enjoys a clear first place while the Barracuda 7200.12 ST31000528AS and the WD Caviars go neck and neck.

Web Server Pattern

Once again the Barracuda LP ST1000DL002 falls behind. The RE4 WD1003FBYX is on top, followed by the Caviar Black WD1002FAEX.

The graphs diverge when the queue is increased up to 256 requests. The Barracuda 7200.12 ST31000524AS and Barracuda LP ST1000DL002 do not accelerate anymore. The RE4 WD1003FBYX and Caviar Black WD1002FAEX are unrivalled. The rest of the HDDs deliver similar performance.

 
Pages: [ 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 ]

Discussion

Comments currently: 4
Discussion started: 11/24/11 10:42:09 AM
Latest comment: 08/26/12 06:19:04 AM

View comments

Add your Comment