What is a fair hard drives lifespan

ballsystemlord · 03-21-2015, 02:20 PM

Alas, my old hard drive, after 8244 hours of faithful service, appears to be dying. The sectors in need of reallocation, have increased rapidly in the last 200 hours of use (it's the spinning type):

Code:

SMART Attributes Data Structure revision number: 16
Vendor Specific SMART Attributes with Thresholds:
ID# ATTRIBUTE_NAME          FLAG     VALUE WORST THRESH TYPE      UPDATED  WHEN_FAILED RAW_VALUE
  1 Raw_Read_Error_Rate     0x000b   071   071   016    Pre-fail  Always       -       242876959
  2 Throughput_Performance  0x0005   134   134   054    Pre-fail  Offline      -       87
  3 Spin_Up_Time            0x0007   172   172   024    Pre-fail  Always       -       307 (Average 361)
  4 Start_Stop_Count        0x0012   098   098   000    Old_age   Always       -       11346
  5 Reallocated_Sector_Ct   0x0033   045   045   005    Pre-fail  Always       -       862
  7 Seek_Error_Rate         0x000b   100   100   067    Pre-fail  Always       -       0
  8 Seek_Time_Performance   0x0005   130   130   020    Pre-fail  Offline      -       28
  9 Power_On_Hours          0x0012   099   099   000    Old_age   Always       -       8252
 10 Spin_Retry_Count        0x0013   100   100   060    Pre-fail  Always       -       0
 12 Power_Cycle_Count       0x0032   099   099   000    Old_age   Always       -       5132
192 Power-Off_Retract_Count 0x0032   091   091   000    Old_age   Always       -       11381
193 Load_Cycle_Count        0x0012   091   091   000    Old_age   Always       -       11381
194 Temperature_Celsius     0x0002   181   181   000    Old_age   Always       -       33 (Min/Max 18/44)
196 Reallocated_Event_Count 0x0032   055   055   000    Old_age   Always       -       984
197 Current_Pending_Sector  0x0022   093   093   000    Old_age   Always       -       319
198 Offline_Uncorrectable   0x0008   100   100   000    Old_age   Offline      -       0
199 UDMA_CRC_Error_Count    0x000a   200   200   000    Old_age   Always       -       0

I wanted to know if I got ripped off on this hitachi.
I normally see HDs that are 8+ years old still functioning, and this on is about half that.

TobiSGD · 03-21-2015, 02:24 PM

There is no way to tell how long a disk will last. Hard diak manufacturers usually will state the MTBF, Mean Time Between Failure, but as the name states, this is a statistical measurement and says nothing about a single disk. Yes, 8000 hours is not that much, but things like that can happen with any disk from any vendor, so I wouldn't call that a rip-off, it just shows why having a good backup scheme is a necessity.

business_kid · 03-21-2015, 02:33 PM

"Life" depends on a number of factors: usage; crashes; file system employed; quality of manufacture; quality of parts; vibration when rotating; Size of disks; speed of disk (2.5 disk drives are designed to accept knocks better); even interface (ide drives seems to survive abuse better than sata ones).

atime on ext2/3/4 is capable f writing every 5 seconds. 'relatime' option cuts that to 15 seconds. Noatime cuts out these extra writes at the expense of file access information. And there are loads of these little eccentricities make your question impossible to answer. I agree your drive is Kaput. Do you move the box while the platter is still rotating? Don't.

jefro · 03-21-2015, 02:39 PM

Heat too might be an issue as well as below temps.

I'd get the Hitachi/IBM drive diags. It may offer better results.

At one time a low level format seemed to help. It may still help.

Plenty of other issues from nearby radar to large motors to power drops/hits to other emf/rfi.

ballsystemlord · 03-21-2015, 02:47 PM

Quote:

Originally Posted by business_kid

atime on ext2/3/4 is capable f writing every 5 seconds. 'relatime' option cuts that to 15 seconds. Noatime cuts out these extra writes at the expense of file access information. And there are loads of these little eccentricities make your question impossible to answer. I agree your drive is Kaput. Do you move the box while the platter is still rotating? Don't.

I always thought that atime was necessary for correct running of your root fs. Some tools, for example fancy du tools, rely on it.

ballsystemlord · 03-21-2015, 02:51 PM

Quote:

Originally Posted by jefro

Heat too might be an issue as well as below temps.
I'd get the Hitachi/IBM drive diags. It may offer better results.
At one time a low level format seemed to help. It may still help.
Plenty of other issues from nearby radar to large motors to power drops/hits to other emf/rfi.

How do I get the diags? I used smartctl to get the partailly shown results above.
When you say "a low level format" do you mean dd if=/dev/zero of=/dev/sda?
Heat, you might be right, I do live in FL USA. emf might also be a factor.

rknichols · 03-21-2015, 06:05 PM

Quote:

Originally Posted by ballsystemlord

I always thought that atime was necessary for correct running of your root fs. Some tools, for example fancy du tools, rely on it.

GUI file managers have pretty much rendered atime a useless statistic in many cases. In order to show an appropriate icon for each file they read the beginning of the file to determine its type. That's going to cause an atime update for every file in the directory just because someone looked in the directory, perhaps just to see what huge files might be there. And, if the file manager resets the atime to avoid the appearance of access, that is going to cause a ctime update and make backup programs think that something about the file has changed.

metaschima · 03-21-2015, 07:41 PM

None of the attributes are failing. Try running a SMART long test, wait for it to finish and post the results. This checks for bad blocks.

jefro · 03-21-2015, 07:47 PM

Low level puts new track info on it. dd just writes over the existing set.

Go to Hitachi web site for diag suite. I think ultimate boot cd has it too.

business_kid · 03-22-2015, 03:47 AM

Quote:

Originally Posted by ballsystemlord

I always thought that atime was necessary for correct running of your root fs. Some tools, for example fancy du tools, rely on it.

I'm not saying it isn't. I simply answered your question. I know one can mount with the -relatime option, I don't know about 'noatime.'

Mr. Alex · 03-22-2015, 04:28 AM

Lifetime of HDDs might vary a lot, depending on many factors. I had ATA Maxtor 40 GB drive working every day, which started to fail after about 3 years of service. And right now there is WD Caviar Blue 250 GB in my Miditower working every day since it was bought in 2008 -- not a single hint on failing. All parameters are like in brand new drive. But back then I had nasty PSU and didn't know about how to handle drives properly. Now I do.

Soadyheid · 03-23-2015, 07:10 AM

Quote:

I wanted to know if I got ripped off on this hitachi.
I normally see HDs that are 8+ years old still functioning, and this on is about half that.

We used to have about 90 x 2.Gb or 4.3Gb Seagate SCSI drives spinning in an old Sun SparcServer 1000e (3 x SSA100 arrays) they'd been spinning for ten years at least.
Latterly, the server sounded like it was a buzz saw. While the disks were still spinning, the bearings had all dried out. You had to spin down a minimum of ten disks to replace one, it was a nightmare! We'd have to "bean shake " the other nine to get them to break the stiction and spin up if we were lucky. The server lasted far too long due to the need for some legacy app or other.

The message being, if they're left spinning, they'll last longer. Starting and stopping them shortens their life.

Play Bonny!

metaschima · 03-23-2015, 10:06 AM

Quote:

Originally Posted by Soadyheid

The message being, if they're left spinning, they'll last longer. Starting and stopping them shortens their life.

Play Bonny!

I agree, that's why I always turn off the spin down power saving feature on laptops HDDs.

ballsystemlord · 03-30-2015, 12:58 PM

Quote:

Originally Posted by Soadyheid

The message being, if they're left spinning, they'll last longer. Starting and stopping them shortens their life.

The disks that I mentioned lasting 8 years are ones that ran windowz and, as a rule of thumb, you start them up and shut them down, preferably daily.

jefro · 03-30-2015, 03:05 PM

I use linux on many computers at work. I haven't noticed any difference between any OS.