CrystalDiskInfo reporting "Caution" on BackUP HDD. Time to worry?

D

doug_7506

2[H]4U
Joined
Oct 17, 2004
Messages
3,247
Just picked up a 3TB HDD from a fellow [H] member. Bought it for the sole purpose of backing up my RAID 1 arrary once a month and to make sure whenever I transfer data or reconfigure my RAID array that I'll have a copy in case stuff goes south. So i'll power it up <20 times a year.

This is the CrystalDiskInfo sheet for it:
jokRQ1j.png


So should I worry with it being in "Caution"?
 
According to cloud backup storage provider BackBlaze, Reallocated Sector Count is one of the five relevant S.M.A.R.T. measurements that highly correlates with hard drive failure. I wouldn't trust the drive anymore.
 
Formatted it with windows quick format and nothing changed.
Click to expand...

That does not help. You need a full format which writes 0s to every sector. That way you can tell if the drive can at least write correctly. If after a full format the # of reallocated sectors grows. Repeat the full format. If it grows again. Time to RMA the drive.
 
drescherjm said:
That does not help. You need a full format which writes 0s to every sector. That way you can tell if the drive can at least write correctly. If after a full format the # of sectors grows. Repeat the full format. If it grows again. Time to RMA the drive.
Click to expand...

Thanks for the help.

Would a format through Disk Management do the job? Or should I use a specific program?
 
Whatever it shows later, I wouldn't trust the drive with anything important. Use it, but treat it as it will die the next second (though this is valid for any drive :)).
 
Meeho said:
Whatever it shows later, I wouldn't trust the drive with anything important. Use it, but treat it as it will die the next second (though this is valid for any drive :)).
Click to expand...

Yeah i'm kinda paranoid about that with my videos and pictures of my kids/family.

Seller is being really cool about it and has offered to swap it. :)

I use 2x3tb in RAID 1. Then 1tb of the important stuff back ups to one drive. Now I want to add a hardcopy backup also.
 
Meeho said:
Whatever it shows later, I wouldn't trust the drive with anything important. Use it, but treat it as it will die the next second (though this is valid for any drive :)).
Click to expand...

I have drives at work that have run 5+ years with a non zero reallocated sectors count. I am only worried if the count grows over time. The point is to find out of the reallocated sectors are from an isolated media defect or some type of failure with the heads.
 
In case you're wondering, 1B is hexidecimal for 27. That's a lot of reallocated sectors. If it increases after the full format, I would retire it immediately.
 
You have 1B realocated but 1C realocation events. That means another sector will soon be gone.

Your UDMA CRC error rate is high, that might suggest a faulty SATA cable. Change it to another one and firmly re-seat it.

I've seen hard drives fail tests and appear broken, with reallocated sectors growing and magically go back to normal after re-seating the sata data cable.
 
I've got a few drives that have had some reallocated sectors for years and they still work fine.

Drives have spare sectors specifically reserved for reallocation. So some amount of reallocated sectors are actually "planned for" by the drive manufacturer.

I would just worry if the number kept increasing, because eventually you will run out of spare sectors.
 
HammerSandwich said:
Nice catch. Shouldn't we see 1 currently pending sector now?
Click to expand...

Probably will after a power cycle and/or a smart self-test routine :)

But that Ultra Ata CRC thing is #1 to watch right now, check it daily, it should not happen at all.
 
if you want to test, boot with bootable linux rescue CD/flash USB.
and run badblock with Read/Write option. you need to copy all data before running badblock due on badblock will overwrite your drive.

let runs in 2 or 3 phase, at leat 1 day torture...
you would see badblock reports bad block detected or not

once is done in 1 day, check smartcl in cmdline, smartctl -A /dev/sXX
check reallocation sector and current pendong sector are increased or not.
 
You can test a disk with this software. If the number stays stable I'd probably keep using it, but of course you have backups?

Also make sure any backups are real backups, not mirrors. If a file corrupts and you mirror it your backups are corrupt. If you have incremental backups you can recover to a known good point.
 
Thanks for everyone's replies. This is CDInfo after the full reformat:

fc2cwCq.png


Have another one coming in from the seller. So it'll be taken care of.

Anyone recommend a good backup software?? :)

EDIT: looks like i have a pretty decent utility included with Nero 12
 
Last edited:
doug_7506 said:
Anyone recommend a good backup software?? :)
Click to expand...
Use more than 1 method.

I've settled on daily FreeFileSync jobs to duplicate files across multiple PCs. Because it does normal copies, browsing/using/restoring these files requires nothing proprietary & works 100% to any kind of storage FFS can see. So, I can have multiple copies across local & remote NTFS, ZFS shares, etc. on different OSes. Good peace of mind.

In addition, I run full image backups less often, using both the Win7/8.1 image tool AND Acronis.

BOTH types of backups get bounced onto external drives on a ~monthly basis.

Paranoia helps, just in case.
 
doug_7506 said:
This is the CrystalDiskInfo sheet for it:
jokRQ1j.png


So should I worry with it being in "Caution"?
Click to expand...

I have a seagate 2TB hard drive that has like 33000 of those relocated errors as well as a 1.5TB seagate drive with 100 hours on it with like 2 errors. They pass the seagate tests. You could do a full erase and then scan for errors which will map out those sectors and use good spare sectors on the tracks. But that function seems iffy..
 
Meeho said:
Whatever it shows later, I wouldn't trust the drive with anything important. Use it, but treat it as it will die the next second (though this is valid for any drive :)).
Click to expand...

You do realise that just because that number is 0 does not mean there are no bad sectors on the drive right? At the factory they map out bad sectors they find, which is why with HDtune you do a drive speed check, the graph is not a slowely moving curve and instead it spikes up and down a little as it sometimes have to use sector later on since the sector where it was, is supposedly marked bad, The cleaner the graph the higher the quality of the platter.

You could do a scan with HDD speed.. Which actually shows you the time it takes to scan a sector. If too many sectors show large variations then the drive is ready for replacement. It actually shows you the time to read each sector.
 
I did a windows backup software review recently. It's targeted at photographers, but it's broadly applicable. It separates backing up large media files and smaller documents, as I couldn't find one backup program I thought was good at both.

Note that I didn't really cover OS backup software, but I might in future. Macrium Reflect free works well though.
 
Keep in mind - a faulty SATA cable will pollute your SMART stats with 'current pending errors', 'uncorrectable offline', even realocation events. If your Ultra ATA CRC rate is going up, you will probably see bad sectors here and there and then disappearing and reappearing somewhere else.
And then when you finally fix the SATA cable and ULTRA ATA CRC stops going up, a full scan will usually clear the 'pending sector' count and all will be well. I salvaged many a drive that way.
Sata cables without the latches are the worst!
 
You must log in or register to reply here.
Back
Top