SMART Failure Predicted on Hard Disk - HGST HGS545(...) 500Gb

Hi everone!

I'm looking for answer for my problem quite long time now, mostly on my country forums but can't get it, sorry for my bad english if some misspells happens

The problem is:

I've reinstalled system, nothing special, just Win 7 reinstall after more than half year of using it. Installed drivers, antivirus, WinRAR, Firefox, T-mobile app for mobile Internet, standard things.

But after reboot - needed by antivirus I've got such message:

SMART Failure Predicted on Hard Disk (tutaj mój dysk)
WARNING: Immediately back-up your data and replace your hard disk drive. A failure may be imminent. Press F1 to continue.

disk is almost new, bought in february - used lot, well thats a fact, but, laptop wasn't moved - first WD HDD died because of a lot travelling, now I'm mostly in home laptop stands on desk so thats a bit weird.

Thats my laptop specs, bit old but still good hardware

Acer Aspire 6930g
Win7 Ult 32bit
Core2Duo 2,0Ghz
Ram 4Gb
Hdd - now using Hitachi HGST 500Gb
GeForce 9600m GT 512(1Gb)

I've scanned disk with HDDtune, Crystal Disk Info, MHDD too, well seems not that bad after all, but problem egzists... 8 sectors realocated - by CDI, and so (screens from programs attached)




MHDD:

AVG: 84811 kb/s
ACT: 53351 kb/s

< 3ms: 3659993
< 10ms: 169397
< 50ms: 1092
<150ms: 0
<500ms: 0
>500ms: 0

? TIME: none
x UNC: 2
! ABRT: none
S IDNF: none
A AMNF: none
0 TONF: none
* BBK: none


I've tried to remap disk but got stuck at 1,5% (those 2 bad blocks are +/- there)

Victoria for Windows also didn't repaired those errors.

It's more than 2 months with no change, sometimes Windows makes alert - then in SMART 01 - RAW gets red but after reboot it's ok again, well not ok, but caution


Any good advice? It would be aprreciated a lot

 

"8 sectors realocated - by CDI"

The drive is going bad. It could be tomorrow, or a year from now, but when a drive starts getting bad sectors...it is on the verge of dying.

 

Well thats obvious, but I'm looking for a way to restore this disk, to remap it or something so I won't get those errors. It's annoying. 8 realocated (by CDI), how to repair this? 2 bad blocks by MHDD - not even one more and thats 2 months+ now. Disk works fine, no lags, no data loss, so it's weird. Ih, I've also downloaded from Hitachi/WD site WinDFT (health tool) but it doesn't even see my HDD while i click scan it frozes for a minute or so, and then goes nothing :/

 

You can't.

S.M.A.R.T. is part of the HDD firmware. The only firmware flashable drives out there are SSD, AFAIK.

When sectors are marked bad, is is because they are physically damaged areas on the drive.\

Sectors that are repairable are due to filesystem damage, which a chkdsk or <insert multitudes of other utilities> would repair, not mark as bad.

 

Looking for second doctor opinion then

Seems unlikely that such error will kill the disk, but I'm no specialist...

next info added: chkdsk finds NO errors...

 

Casemiro, please don't take offense, but the idea of trying to "save" a drive that is failing is penny wise and pound foolish.

Your data and time are worth far more than the cost of replacing the drive. Since is has not failed completely, you may be able to "clone" the old drive to the new one.

Since the current drive is under one year old, it should still be under warranty. Go to Hitachi's website and check the serial # in the warranty support option. They may have the option of "advance replacement," were they send you a new or refurbished drive in advance and guarantee you'll return the old one with a deposit charged to your credit or debit card.

Hope this helps.

 

chkdsk finds no errors because CDI, as you said, marked the sectors as bad. As such, they are now not considered usable, and are moved to the end of the drive and ignored.


The error isn't killing the disk. It is an indicator of disk failure. It is an effect, not a cause.

 

If you don't have work it's not stupid but wise to try to repair rather than spend money on new. Warranty, yes of course but if they won't send me new in replacement it could be couple of weeks without only one working computer in my home, and thats a pain for me, especially if I'm looking for a job now, mostly on Internet sites, but thats enough. If ya can't help so be it. Hitachi here I come ;]

 

To repair a hard drive costs more than to buy another. It takes expensive, specialized equipment.

 

Then you got it. Hard drives are like metal rods. If the rod is bent just once, you can re-straighten it, but it will NEVER be as strong as it was before, and will likely bend again soon, and eventually break totally - perhaps catastrophically.

IF the damage to the hard drive was ONLY corrupted data, then, and only then, is the drive safe to keep using to store your precious data.

IF the errors are due to a fault in the SMART monitoring coding, then you might be safe to disable SMART monitoring and store your data there.

But those are "big ifs".

The errors you got indicate much more than simple corrupt data. It is good that chkdsk did not find any errors, and you may be able to simply disable SMART monitoring and keep using that drive for years. But IMO, you would be skating on thin ice and you should take that error message as a sign of impending doom! You should (1) replace the drive, or at the very least, (2) create and maintain current backups of the data on that drive, and be ready to replace the drive and restore from those backups.

If hard disk space was as expensive as it was 10 - 15 years ago, I would be more sympathetic to your desire to keep it running. But with today's monster drives costing less than $.04 per gigabyte, it does not make sense to put or keep your data at potential risk 24/7.

The problem is trust and integrity of YOUR DATA. It is not likely to "kill the disk" - just render it unreliable.

 

Thanks I wrote to Hitachi, they'll do it on warranty. But as I said it'll take time about prices - 200PLN+ in country where most people get 1300PLN/month is not small amount, especially when you are not employed anyway thanks

Resolved

 

I hear you when it comes prices. Being retired military - no way am I flush with funds - I am on a restricted budget too. But when it comes to hard drives, you need to look at what it would cost you to replace all your lost data too. And you need to look at the cost of lost productivity while the system is "down", and the value of your time that you spend recovering from such a loss.

For many, including myself, my data is worth MUCH than all the computers, printers, and network gear in my house.

Anyway, great news that Hitachi will replace the drive under warranty.

 

If you still need to address this problem, here's a workaround I've used for NON-CRITICAL data only. It looks complicated, but in reality it's just a few simple steps...

Isolate the bad blocks to a partition (or partitions) as needed that you will NEVER use for any purpose. This will NOT stop the SMART errors or possible predicted time-to-failure dates, but will allow you to use the drive with substantially reduced fear that your data or system will be corrupted.

1) Use the bad block info in the HDtune "Error Scan" to determine approximately where on the HDD the bad blocks physically are located.

2) Then, format the HDD into multiple partitions (note: if a boot drive, this workaround will not work if the bad blocks are located in the MBR or Boot partition space on the drive and would include the bad blocks). But, keep in mind that some common (e.g. My Documents), temporary, and profiles/settings files can be relocated to different partitions and/or different HDDs entirely, which could substantiallydecrease the needed size of the boot partition.

For example,
(assumes bad blocks:
a) on two widely separated physical regions of the drive
b) were identified by HDtune at about 133GB on the drive and about 205GB)
So, I might format and partition the drive as follows:
Partition 1=0 to 132GB (boot partition, including any non-relocated tempory/configuration/settings space)
Partition 2=132GB to 135GB (and includes the physical locations of the 1st region of bad blocks)
Partition 3=135GB to 200GB
Partition 4=200GB to 210GB (and includes the physical locations of the 2nd region of bad blocks)
Partition 5=210GB to end of drive

3) Then, if desired, make Partition 2 and 4 read only so you don't accidentally write to it or store critical data on those partitions.

NOTES:
This does NOT eliminate all chance of future bad block problems and this definitely doesn't cover all the bases, but bad blocks are often located in more or less proximate physical locations on a drive. As the drive ages, there may be future bad blocks detected by HDtune, but they will LIKELY (but not always) be proximate the previously identified physical positions of any current bad blocks. If desired, you can increase the size of the "bad block partition(s)" to further reduce the chance of future bad block "growth" impacting data integrity, by keeping the identified bad blocks in the middle of the partition. In the procedure above, you might decrease the size of Partition1 and Partition3 to make partition2 larger (for example, from 120GB to 145GB and adjust the other partitions accordingly). However, if you have a drive that has a lot of widely physically separated bad blocks, I'd scrap/recycle the drive and/or scavenge it for interesting parts (powerful magnets, platters, servos, etc)

Internal drive remapping should not affect the location of the bad blocks as reported by HDtune.

Backups and other utilities: You should also note that some disk/drive utilities simply will not work on a partition/drive in which a bad block exists. A given utility may for example, simply cease operations and post an "unrecoverable error". Thus, you may be able selectively to backup a good partition, but be unable to image the entire drive itself.

Not suitable for drives used in either software or hardware RAID systems.


To "eliminate" the errors reported by SMART:
You MAY be able to use HDD Scan's "erase" function on the drive. This MAY temporarily "reset" the identified bad blocks. It worked on my drive. Then format/partition the drive. But IMMEDIATELY after mark the bad block containing partition(s) as read only or make it(them) otherwise inaccessible to users. Note: bad block errors MAY come back in the future. Keep in mind that even though the "bad blocks" LOOK like they were "fixed"...they were not actually fixed. I do NOT recommend using for data/OS any region on a drive in which bad blocks ever have been identified.

 

As the drive is still less than a year old, then probably you could still claim for a warranty. Best thing you could do is to back up your important data and ask for a new replacement drive. Good Luck

 

A few posts down from first post: