Uh, Oh

[Mitchum22]

Something strange happened over the weekend. Did a harmless reboot, and for the first time ever I received the following message before anything loaded:

"A hard drive problem has been detected. Consult your troubleshooting guide for assistance.

Please press [ENTER] to continue."

Which I did, with no problem. I've since rebooted several times, with the same start up message. I've done a CHKDSK(with no problems) and a defrag using DiscKeeper.

What's happening here? I know it ain't good.

[BJ_M]

prob got a bad sector .. you have to purge the HD log to get rid of the msg ... or turn off HD monitoring ..

[Mitchum22]

Dear Moderator,

Thank you very much.

How do I purge the log, or turn off HD monitoring?

Is it possible the HD's about to fail? Everything seems fine.

[jagabo]

Is it possible the HD's about to fail?

Yes.

When I start seeing SMART warnings like that I get a new drive ASAP. I then copy all the contents to the new drive and set the old drive aside for a while (as a backup). If everything goes well with the new drive I'll eventually reformat the old one and use it for storage of unimportant data. If it eventually fails it's no loss. If it doesn't fail it's extra storage space.

[ROF]

That is a message coming from the S.M.A.R.T. HDD technology. Your hard drive is more than likely going to be failing very soon. This is a warning.

[Mitchum22]

Jagabo and ROF,

Thanks.

I guess it's back-up and replacement time. . .

Since I just received this, how long would you guys say I safely have 'til a crash???

[ROF]

1 hour to 10 years . . . too hard to predict.

[greymalkin]

1 hour to 10 years . . . too hard to predict.

yep..don't risk it man. Backup immediately if you haven't already done so.

[Mitchum22]

All,

Thanks for all the help.

Not going to risk it. It's going in tonight.

Just out of curiousity, I've now used two Hard Drive Heath programs:

http://www.panterasoft.com/pics/hhealth.gif

http://www.hddlife.com/

And they both indicated all was OK. Hmmmmm....

Not risking it.

[b1tchm4gn3t]

I know I got a SMART error one time and it said to 'backup my data'. I wish I could have, the drive no longer worked immediatly after that message. So my advice, backup asap!

[Mitchum22]

Uggg.

That's clinched it.

[kschang]

Use this one to pull the SMART registers off the HD directly:

http://www.passmark.com/products/diskcheckup.htm

Free for personal use.

[jagabo]

Use this one to pull the SMART registers off the HD directly:

http://www.passmark.com/products/diskcheckup.htm

Free for personal use.

Nice little utility!

[Mitchum22]

kschang,

Thanks for pointing me to the utility.

The failure I'm getting, on my Maxtor 6Y080L0 HD, is "Reallocated Sector Count":

RawValue: 2090
Value: 47
Worst: 47
Threshold: 63

I have no idea what any of this means.

[jagabo]

When a sector (usually a 512 byte block) goes bad on a modern drive, a good one is automatically substituted from a pool of spares. So some parts of your drive have developed bad sectors (it's not unusual to have a few) and you are getting a warning indicating as much. There's no way of telling if the problem is going to get worse.

[TBoneit]

Just for a point of reference I looked at my three drives and two showed a reallocated sector count of 0 (Zero) and one showed a count of 1 (One).

Once you get the S.M.A.R.T. error like that it is only a matter of time before the drive dies. It can be the next time or a while...

The only sure thing is that the drive is dying when you get that error. If loosing everything is not a matter of importance then ignore it. But if it is your boot drive have a spare on hand and be prepared to reload.

[BJ_M]

jagabo is right - a lot of drives (even new ones) may have a few bad sectors... its when (as in your case) they start going out of the blue that bad things will happen ...

at least you dont have it clunking away in vain yet - i really hate that sound and makes me kind of sick ...

[kschang]

kschang,

Thanks for pointing me to the utility.

The failure I'm getting, on my Maxtor 6Y080L0 HD, is "Reallocated Sector Count":

RawValue: 2090
Value: 47
Worst: 47
Threshold: 63

I have no idea what any of this means.

RawValue doesn't mean anything to you. The important is Value and Threshold.

Value = current value of relocated sectors
Threshold = the do-not-exceed value.

Most HD's should only have a few, 0 or 1 or maybe a handful. You're already at 47. I'd keep an eye on it to see if that number is still growing by the day or week.

The reallocated sector is basically an "error avoidance" by the HD. THere's an area on the HD that's reserved for the system. If it detects any errors, it automatically subtitutes a sector in this "spare pool" and makes a note in internal cache to do this automatically in case that sector is referenced again. It's like having mail forwarding set. Any request for that sector is handled dynamically by the HD, the O/S sees nothing. However, SMART keeps track of these sector failures, and you're getting close to the "darn it" level.

[Mitchum22]

kschang,

Thanks again for the great help.

And thanks to the utility you pointed me to, I've been able to monitor things for the past few days. The 47 value has remained at 47. ('Though I'm still getting the "a problem with your HD etc" message on boot-up.)

I did get a new WD hard drive 3 times the size of the old one. So if the ghosting goes okay, I'll be better off than before.

Thanks again!

[Seeker47]

Once you get the S.M.A.R.T. error like that it is only a matter of time before the drive dies. It can be the next time or a while...

The only sure thing is that the drive is dying when you get that error. If loosing everything is not a matter of importance then ignore it. But if it is your boot drive have a spare on hand and be prepared to reload.

Unless we're talking SATA, or the 500G + variety, hard drives are fairly cheap these days, especially given the regular sales and rebates. Why go through all the grief of reloading and rebuilding everything ? It is a much better idea to clone your boot drive, and keep that spare as up to date as possible. (You can also make partition image "snapshots" with appropriate software -- in-between whole drive clonings -- and dump them off to other storage. You can keep some of that other storage off-site too, just in case.) That should provide good protection against a catastrophic failure, and losing everything.

[Mitchum22]

Just a little follow-up. Checked the HD again with kschang's recommended program and the number of errors has gone down from 47 to 45.

Hmmm.

[Poppa_Meth]

Did you try scanning it with the HDDs maker's utility? I've gotten several errors on a maxtor drive before that chkdsk couldn't even see (acronis caught them) and the manufacturer's utility found them and managed to repair them.