11

What are the main differences between desktop-series hard disks and server-series?

The obvious things I can see are: durability (server hardware mostly more qualitative and have more warranty) and power consumption (server hardware more focused on performance, than on power economy). Also server disks are usually a little faster, but it seems, that it is not always the case.

May be there are some other reasons, that make you choose server-oriented series (Seagate ES drives, for example) over desktop-oriented ones (Seagate Barracuda series)? What are they?

Vladislav Rastrusny
  • 2,581
  • 12
  • 39
  • 56

6 Answers6

9

One very important difference is the Time-Limited Error Recovery (aka Command Completion Time Limit)

Time-Limited Error Recovery (TLER) is a name used by Western Digital for a hard drive feature that allows improved error handling in a RAID environment. In some cases, there is a conflict as to whether error handling should be undertaken by the hard drive or by the RAID controller, which leads to drives being marked as unusable and significant performance degradation, when this could otherwise have been avoided. Similar technologies are called Error Recovery Control (ERC), used by competitor Seagate, and Command Completion Time Limit (CCTL), used by Samsung and Hitachi.

This is very important in RAID arrays where one drive can lock up or degrade the array.

tegbains
  • 1,956
  • 12
  • 27
  • 5
    it is a good idea to look for "RAID-compatible" drives, since most consumer drives will never be used this way. – Jeff Atwood Apr 30 '10 at 11:36
  • A late response, but I have to reply. I had three `WDC WD2002FYPS-02W3` disks (with TLER) fail on me in a couple of weeks and they didn't limit their error recovery time at all. The Linux MD driver never knew the drives were broken. **I** knew because the server's load was getting high and dmesg was full of errors. – Halfgaar Jul 11 '13 at 14:09
9

According to Intel's Enterprise-class versus Desktop-class Hard Drives, enterprise-class drives are often faster and more reliable due to better hardware and different firmware.

Better hardware specs:

  • better mechanics for faster and more reliable data access (stronger actuator magnets, more servos)
  • more cache memory
  • more components for error detection and correction
  • vibration compensation or reduction to reduce likelihood of data corruption induced by moving parts in the server, e.g. rotating fans and spinning disks

Behavior:

  • Time-limited error recovery (TLER) for more reliable and lower-latency error recovery and fail-over
  • End-to-end error detection
  • Usually fixed rotation speed
  • Energy-saving mechanisms that shorten overall life of consumer-grade disks, such as Western Digital's WD Green series' IntelliPark which parks the hard disk after 8 seconds (!) idle time (however, there are tools from the manufacturers and third parties that modify the default settings so that you could use such disks in server or NAS/SAN systems, e.g. idle3-tools).

Testing:

Warranty:

  • Enterprise-class disks often come with longer warranty than desktop disks. However, this will probably not matter very much to you if you have sensitive data stored unencrypted on your failing disk and do not want to send in your disk to claim the warranty.

Nonetheless, the issue about TLER might be the main reason for choosing server-grade storage solutions over consumer products, especially when operating RAID systems and servers with time-critical workloads like web or database servers.

Other than that, yes, there is probably some FUD by the manufacturers to make you feel uneasy about using desktop products for a server, so using enterprise products will also give you some peace of mind.

Arc
  • 343
  • 4
  • 9
  • Wondering what the manufacturers' policies about planned obsolescence are regarding their enterprise and consumer products. Might go either way - businesses are more willing to replace hardware and tend to design for redundancy anyways, so why not have enterprise hardware fail more often ? – Arc Jan 09 '14 at 15:36
5

THE most important difference is the level of Vendor Support.

If you don't use "hardware compatible" equipment (i.e. OEM) then you may find yourself SOL when the kit dies and you need to get it fixed quick.

Guy
  • 2,658
  • 2
  • 20
  • 24
  • How vendor can support hard drive other than exchange it on warranty? Do HDDs require any support from vendor after they are purchased? – Vladislav Rastrusny Apr 30 '10 at 11:27
  • 1
    @fractal not really, but the vendors tend to be jerks about this if they can.. if you put in a "non-authorized" drive they'll refuse to service the machine. – Jeff Atwood Apr 30 '10 at 11:35
5

Enterprise hard disks also tend to have a longer MTBF (Mean Time Between Failures), i.e. longer life span, than consumer grade ones.

sheepbrew
  • 227
  • 2
  • 3
4

The heat-pad under a smooth chip (motor controller) is missing in the most of the consumer drives (seagate, WD), but is usually present in the most of the enterprise class disks. (so the tiny smooth chip is much more likely to die under the load from overheat in the desktop model). Also enterprise drives usually have a couple of the vibration sensors. Rubber anti vibration dumpers and rigid cases are strongly recommended for all modern high density disks/fan mounts. (Esp ones over with density 500GB+ per platter, since their heads are really fragile from my experience).

Mark
  • 41
  • 1
0

Two things:

  • Cache size
  • Marketing

Reliability is not a distinguishing factor.

stribb
  • 59
  • 1
  • 1
    agree, it's 95% marketing. There might be longer warranties if you pay more for the "enterprise" stuff, but economies of scale dictate that the drives will be basically identical. – Jeff Atwood Apr 30 '10 at 11:35
  • Some desktop models also have 32-64Mb cache size: http://www.seagate.com/www/en-us/products/internal-storage/ – Vladislav Rastrusny Apr 30 '10 at 11:39