Reallocated sector count 2752835 on HTS541680J9S(80gb)

[dhiman33]

OMG my hitachi laptop hdd which is almost 4 yrs old shows reallocated sector count attribute as 100:100:5:2752835, reallocated event count 100:100:0:39 and also ultra dma crc error count as 200:253:0:3829---all these warnings according to hdtune pro 4.60. Everest also reports same things but shows only the reallocated sector count as warning.
For many months the reallocated sector count stayed somewhere between 2550000 to 2560000(can't remember exact value). Then a few days ago I checked it and found the value between 2650000 to 2660000. Now today its 2752835.Hdtune reports each sector size to be 512 bytes. That means almost 1.5 gb sectors are damaged!

Still, the hard disk gives good performance and everytime passes smart test--- I've also tested with Hitachi Drive Fitness Test and everytime it passes without problem.

I think the number increased because there were frequent power outages these days and since my laptop battery does not function so the laptop turned off 2-3 times.(write caching is disabled for windows, still it can hurt a lot?)

O.k. so now I ask what should I do? Personally I prefer waiting until smart check fails and then buying a next hdd(and in the meantime, backing up everything) Do u agree? Also, which hard disk would u advise to buy then?(within 160 gb since I'm on a low budget). And what do u think about the remaining life of the disk, can it run 4-5 months more?

[fzabkar]

The normalised value is still at 100, so it's nowhere near as bad as you think.

You need to examine the data in hexadecimal format.

Post a HD Sentinel or HDDScan SMART report.

http://www.hdsentinel.com/
http://hddscan.com/

BTW, HD Tune only reports the lower 32 bits of each raw attribute value. The author doesn't appear to understand that the data span 48 bits, and that important information sometimes resides in the uppermost 16 bits.

[dhiman33]

Thanks for replying,fzabkar. I'm attaching report of hddscan. Hdsentinel shows my performance is 100%(Excellent), health is 17%(Bad), it says I have 323 remapped sectors,1 weak sector(formed yesterday)...and recommends me to immediately backup everything. Power on time is shown as 339 days 4hrs, and estimated remaining lifetime is 42 days. Now what to do? Is the hard disk that seriously ill?

Attachments

Desktop.7z

(35.09 KiB) Downloaded 621 times

[fzabkar]

According to HDDScan, the Reallocation (sic) Sector Count is 0x0000002A0143.

HD Tune's author converts the entire value into a reallocated sector count, ie 2,752,835.

http://www.google.com/search?q=0x000000 ... in+decimal

HD Sentinel's author understands that this number makes no sense, so he uses the lowest 16 bits, giving a count of 323.

http://www.google.com/search?q=0x0143+in+decimal

IMO this still looks too high. Hard drive manufacturers allow for around 2000 reallocations before a drive is considered to have failed SMART. If the figure of 323 reallocations were correct, then there should have been a significant drop in the normalised value. Instead the normalised attribute value remains at 100.

I'm guessing that the actual number of reallocated sectors is probably 0x002A, which corresponds to 42. However, I'm still not comfortable with this figure, as I would have expected a loss of at least 1 point from the normalised value. I notice that there is one pending sector. It would be a good idea to monitor the Reallocated Sector Count and note whether this pending sector, if and when it is reallocated, increments the lower 16 bits of the raw value, or the middle 16 bits.

In summary, I'd say the drive is sick, and that a backup would be advisable. However, it isn't anywhere near as bad as either HD Sentinel or HD Tune make out.

[dhiman33]

So its in much better condition than the reports? So how to know exactly how many sectors are remapped or how healthy the disk is? Should I post the result here once the reallocated sector count increases again?

Also can u somehow predict how many approx. days/months this hd may go(if not 42 days as reported by hdsentinel). I am very confused right now--- is it o.k to use this hard disk in usual manner? Can u tell me what average lifetime these types of laptop hard disks have(mine is 339 days 6 hrs).Also can I wait until smart fails---then backup everything(I've got an exam soon)---the hard disk should work o.k for say 1 week after smart fails--isn't it?

And once the smart fails...what should I do with the hd? Is there a way to use its still good sectors to make it an external backup device to store movies,games etc?

[fzabkar]

Backup the data that you want to keep. As for predicting drive failure, if a doctor gives a terminally ill patient 6 months to live, you wouldn't expect that he will live for 6 months to the day. Instead, you wouldn't be unduly surprised if he lasted a week or several years.

I can't answer your question as to how many reallocated sectors your drive really has. SMART data are not standardised between manufacturers, or even between models from the same manufacturer. Furthermore, manufacturers don't make their SMART documentation publicly available, so all you can do is to make educated guesses, or decipher the meanings of each attribute by analysing SMART reports from several drives, preferably failing ones.

[Vulcan]

I'm guessing that the actual number of reallocated sectors is probably 0x002A, which corresponds to 42.

Agreed. Based on my experience with these & similar Hitachi 2.5" disks, it's the middle 4 nibbles (16 bits) of the raw value which holds the count of reallocated sectors (so 0x002A in this case). The value in the lower 4 nibbles can vary widely (even for small counts of realllocated sectors) and I have not been able to find what measurement the lower 4 nibbles is showing

@dhiman33 - You've asked many questions which I don't have time to answer in detail. You seem to be expecting too much from SMART - as has already been explained to you, this is not an exact science. I will answer one of your questions:

the hard disk should work o.k for say 1 week after smart fails--isn't it?

No, it is not that simple. Disks can suddenly become unreadable (or have any other kind of abnormal behaviour) at any time (e.g. before, or immediately after, or a long time after any SMART warning). Therefore my advice is that you should not rely on the drive working correctly for 1 week after it reports a SMART predictive failure, especially since we have evidence that the disk already has some reallocated sectors. Do you feel lucky?

[dhiman33]

Hmmm...these answers certainly reduced my panic a lot. Thnx for answering, both of u(And yes,I do feel lucky )

[fzabkar]

My curiosity was killing me, so I looked around the Internet for other SMART reports. However, I couldn't find a single one where the normalised value had budged from 100.

https://bugs.launchpad.net/ubuntu/+sour ... bug/438136

"I get a reallocated sector count of 65551 [= 0x0001000F] on my Hitachi HTS541680J9SA00, very similar to the value 65543 [= 0x00010007] reported by Tapas Bose."

"My reallocation sector count [for a Hitachi drive] is 65538 [= 0x00010002], and the rest seems unremarkable to me (only reallocation event count is 1)."

http://forum.hotfix.pl/problemy/bsod-10 ... t8022.html
(05) Reallocated Sector Count 100 100 5 0 ok
(C4) Reallocated Event Count 100 100 0 58 warning
(C5) Current Pending Sector 100 100 0 0 ok

http://sadmin.org/forum/23-625-1
5 Reallocated sector count 100 100 5 3
196 Reallocated event count 100 100 0 2
197 Current pending sectors 100 100 0 0

http://launchpadlibrarian.net/34574037/smartctl.txt
5 Reallocated_Sector_Ct 0x0033 100 100 005 Pre-fail Always - 327697 [= 0x00050011]
196 Reallocated_Event_Count 0x0032 100 100 000 Old_age Always - 4
197 Current_Pending_Sector 0x0022 100 100 000 Old_age Always - 0

http://launchpadlibrarian.net/35971054/ ... _tests.log
5 Reallocated_Sector_Ct 0x0033 100 100 005 Pre-fail Always - 65542 [= 0x00010006]
196 Reallocated_Event_Count 0x0032 100 100 000 Old_age Always - 1
197 Current_Pending_Sector 0x0032 100 100 000 Old_age Always - 0

http://launchpadlibrarian.net/36599746/ ... s-deer.log
5 Reallocated_Sector_Ct 0x0033 100 100 005 Pre-fail Always - 65552 [= 0x00010010]
196 Reallocated_Event_Count 0x0032 100 100 000 Old_age Always - 1
197 Current_Pending_Sector 0x0032 100 100 000 Old_age Always - 0

https://bugzilla.redhat.com/attachment.cgi?id=382378
5 Reallocated_Sector_Ct 0x0033 100 100 005 Pre-fail Always - 655424 [= 0x000A0040]
196 Reallocated_Event_Count 0x0032 100 100 000 Old_age Always - 10
197 Current_Pending_Sector 0x0022 100 100 000 Old_age Always - 0

https://bugzilla.redhat.com/show_bug.cgi?id=506254
reallocated-sector-count 100/100/ 5 FAIL 1900724 sectors Prefail [= 0x001D00B4]
reallocated-event-count 100/100/ 0 n/a 22 Old-age Online
current-pending-sector 100/100/ 0 n/a 0 sectors Old-age Online

http://www.elektroda.pl/rtvforum/topic1 ... ml#7012633
(05) Reallocated Sector Count 100 100 5 5 ok
(C4) Reallocated Event Count 100 100 0 6 ok
(C5) Current Pending Sector 100 100 0 0 ok

http://forum.kooora.com/f.aspx?t=8921669
05 Reallocated Sector Count 5 100 100 0
C4 Reallocation Event Count 0 100 100 43
C5 Current Pending Sector Count 0 100 100 0

[Vulcan]

However, I couldn't find a single one where the normalised value had budged from 100.

Given the relatively low raw Reallocated Sector counts seen on those web pages (highest seemed to be 58?) then those examples fit with my experience too.

The last time I can remember seeing a significant change in the normalised ("cooked") value for that parameter, was on Segate Moose drives with >1000 reallocated sectors. However, for some modes of failure the drive can become unusable (at least for a normal end user) long before that many sectors have been reallocated, and hence (even assuming that SMART monitoring software is being run) long before a SMART predictive failure warning is seen

Which is why, of course, users must not rely on seeing a SMART warning before starting to do backups

[Vulcan]

these answers certainly reduced my panic a lot.

IMHO you should still be concerned, if you care about the data on that disk.

Even though the number of reallocated sectors seems small (compared to what you thought the number was at first!), the fact that this number is non-zero means that statistically this drive is much more likely to fail, and you may get no warning of that failure.

[drc]

Even though the number of reallocated sectors seems small (compared to what you thought the number was at first!), the fact that this number is non-zero means that statistically this drive is failing

IMO

[Vulcan]

@drc: I certainly understand your change to my comment, and I would always treat a drive having non-zero reallocated sector count with suspicion, as I hope I explained

However I've also seen (over many years of examining SMART data) just a few drives where the reallocated sector count did stabilise at a non-zero value, and the drive then continued to be usable for a long period (i.e. years). So while the vast majority of drives I've seen with a non-zero reallocated sector value did prematurely fail (as the Google data centre experiences also show), personally I stop just short of stating definitively that such a drive is failing, as I found that a few would continue to work for an extended period in that state

But that's just based on my experience and is also IMO

[drc]

I guess it depends on our definitions of fail... To me once the drive has bad sectors that are not recoverable by overwriting, it has already demonstrated a failure to do its job correctly and should be replaced.

[Vulcan]

@drc: I certainly understand what you mean

Although I've changed to a slightly different view, after discussions with disk manufacturer's FAEs and reading their failure analysis reports over the years, I'm not saying that I'm "right" (or that you're wrong) - we all have different experiences

[Doomer]

According to HDDScan, the Reallocation (sic) Sector Count is 0x0000002A0143.

According to HDDScan the Reallocation Sector Count is 0x0143

[fzabkar]

The last time I can remember seeing a significant change in the normalised ("cooked") value for that parameter, was on Segate Moose drives with >1000 reallocated sectors.

IME Seagate drives will lose 1 point from the cooked value for every 40 reallocated sectors or thereabouts.

Here are two examples of my own:

http://www.users.on.net/~fzabkar/SmartUDM/13GB.RPT
http://www.users.on.net/~fzabkar/SmartUDM/40GB.RPT

The 13GB drive still works, but it grows new defects on a regular basis. Therefore I don't use it for any critical data.

[Vulcan]

IME Seagate drives will lose 1 point from the cooked value for every 40 reallocated sectors or thereabouts.

I managed to find the SMART data from one of the more faulty 1TB Moose drives I was referring to (I thought I'd deleted this long ago) - it shows the following (which is an even more significant reduction in "cooked" value than I remembered, but still not enough to push it below the threshold, although this is not the best example that I originally had):

Code:

ID# ATTRIBUTE_NAME          FLAG     VALUE WORST THRESH TYPE      UPDATED  WHEN_FAILED RAW_VALUE
  5 Reallocated_Sector_Ct   0x0033   059   059   036    Pre-fail  Always       -       851


Therefore the Moose drives don't appear to follow the same formula as you describe on your (non-Moose) drives - but as we know, the algorithms used to produce the "cooked" values are commercially sensitive (perhaps not linear?) and so we don't know when / why / how they change between different drive families / makes etc. IMHO it's an interesting comparison nevertheless, thanks

[dhiman33]

I must say this forum is very helpful... the way u people are responding is unbelievable.. Ok. now I'm listing things I've understood so far...pls correct me if I'm wrong.

a)S.M.A.R.T. isn't very reliable and my hdd can fail even before S.M.A.R.T fails.
b)My hdd has 42 reallocated sectors, the number is not 323.

If I consider S.M.A.R.T to be reliable, then---

c)Until the normalised value drops from 100, there's nothing to worry whatever the sector count is.(Or does it always stay at 100 for hitachi?)
d)The health of the disk is well above 17% and I can expect it to last way more than 42 days.
e)My hdd has atleast 500 available sectors for reallocation(fzabkar said 2000, lets take it much less than even that).So until the number reaches this limit...the hdd has low probability of failure.

[fzabkar]

c)Until the normalised value drops from 100, there's nothing to worry whatever the sector count is.(Or does it always stay at 100 for hitachi?)
d)The health of the disk is well above 17% and I can expect it to last way more than 42 days.
e)My hdd has atleast 500 available sectors for reallocation(fzabkar said 2000, lets take it much less than even that).So until the number reaches this limit...the hdd has low probability of failure.

(c) I expect that the cooked attribute value will drop to 99 after your drive grows a handful of new defects.

(d) If HD Sentinel's author had used a figure of 42 instead of 323 for the reallocated sector count, then he would have rated your drive's health much higher than a mere 17%.

(e) I don't think you can trust your drive to the extent that you are willing to. It only requires one bad sector in a critical file system area to render your data inaccessible.