How To Interpret S.M.A.R.T data?

F

FenFox

Limp Gawd
Joined
Dec 20, 2016
Messages
296
So I've browsed through numerous articles discussing S.M.A.R.T data, but I'm still not sure how to interpret the data. Can someone give me a bullet point explanation of how it works? This is a S.M.A.R.T scan of a brand new 10 TB WD Red drive.

How To Interpret SMART data.png


Current/Worst/Threshold mean what--in terms that your average user would understand?

I don't know if I'm under or over the "Threshold" etc. I know what the "Attributes" mean from reading about them, just not how to interpret the values provided.
 
I never pay attention to those columns, the raw value shows you what you need to know and that is a healthy drive!
 
PliotronX said:
I never pay attention to those columns, the raw value shows you what you need to know and that is a healthy drive!
Click to expand...

So you're saying you never pay attention to current, worst or threshold?

How are you interpreting the "raw values" to be healthy? Like what looks unhealthy?
 
The raw value is what is being read from the drive's controller - the S.M.A.R.T. subsystem is not the controller, just an additional aspect of the circuitry to monitor (hence the M) the drive's status. My personal recommendation: don't trust S.M.A.R.T. status 99% of the time, the only thing that truly matters in most instances would be using the manufacturer's diagnostic tools to do a check on the drive and if that diagnostic tool says you have a real problem, then yes, you've got a real problem. Over the years I've had drives that had 100% clean S.M.A.R.T. status and the drives were useless; in other instances I've had drives that were perfectly fine with terrible status readings, and I currently have a Hitachi laptop hard drive that has 8 million bad sectors on it according to the S.M.A.R.T. status and yet when I run the Hitachi DFT (Drive Fitness Test) I get zero errors. I run MHDD (a DOS-based hard drive diagnostic tool) and get zero errors. I've run multiple different tests using a wide variety of software and every one of them tells me the drive is fine but that S.M.A.R.T. status says this:

SMART_SUCKS.png


The hex value of 7A03D8 works out to 7,996,376 sectors and yet the drive has no actual errors aside from the defective S.M.A.R.T. subsystem itself. Now, I'm not saying you shouldn't make use of diagnostic tools if you have some actual reason to be concerned or suspect there's an issue with the hard drive or whatever storage media you're using, I'm saying that for myself, I don't trust S.M.A.R.T. any further than I could throw a so-called defective drive I suppose.

Anyway, OP, in your example, the relevant parameter that most people focus on is the 05 parameter, Reallocated Bad Sectors Count, and on that drive your raw count (which is what the reading actually is) is 0 (yes it's 000000000000 in hexadecimal but it's still zero as in there are no sectors that have been remapped aka marked as bad by the drive's controller. The other parameters can show something wrong if something is actually wrong but in your case, that drive is clean as a whistle.

The Threshold variable typically means the S.M.A.R.T. status for that parameter will not be "tripped" or you won't be alerted in some manner that there could be a problem: the Threshold for 05 on your drive is 67 (that's a hexadecimal value which in decimal is 103) so the drive won't consider things wrong until you get at least 104 actual remapped sectors marked as bad, basically.

The Attributes are all the parameters the S.M.A.R.T. subsystem monitors and no, you don't necessarily have to know what they mean, that's not the point of the whole thing and again as I said the one parameter or attribute that matters the most, especially for physical hard drives and not so much for SSDs, is 05 Reallocated Bad Sectors Count. The rest of them are just informational more often than not and not a big focus - if there's an actual problem, you'll know it.

You're fine, the drive is fine, but I've noticed you've been asking a lot of questions about hardware of late and so far whatever machine you bought or built appears to be working fine. Stop worrying so much and put the damned thing to use. :)
 
Tiberian said:
The raw value is what is being read from the drive's controller - the S.M.A.R.T. subsystem is not the controller, just an additional aspect of the circuitry to monitor (hence the M) the drive's status. My personal recommendation: don't trust S.M.A.R.T. status 99% of the time...
Click to expand...

Yes, the "status" is not helpful sometimes. The raw data can be misleading as well since sometimes raw data is published as several values (or rates) crammed into one hex value. For instance, your value of "7A03D8 works out to 7,996,376" it COULD be 7,996,376, or it could be 7:996,376 or 79:96,376 or 799:6,376. To find out you have to look at the published data for the drive, if the manufacturer has released it (which some don't). In the end keeping a drive running is up to user discretion. Personally, if this was my drive I would keep it in service? Probably not, this is a 2.5" drive, if this is a laptop or a portable drive with this many errors it probably has several areas of head crash. I may use it provided it was part of an array with a journaling fs. Would I trust it as the sole copy of data? No.
 
I don't think you read my post carefully: I said I ran the Hitachi DFT on it, both short and the long tests (the short one, 8 minutes, the long one, 4 hours plus), the drive has no errors, period. I then did 2 passes with MHDD (surface scans) and those took nearly 6 hours and again, the drive has no errors, period. The surface of the platters in this hard drive have no defects, there has been no head crash (if that were the cash that would be painfully obvious), and so on.

It blows my mind that modern hard drives can have close to 100 billion sectors and people want to toss 'em in the garbage if their drive shows a handful of bad ones, it's ludicrous. :)

I trust Hitachi drives more than any other brand because not one of them has ever failed on me, nor the older IBM manufactured drives either. I even have a "DeathStar" 60GXP that I bought in 1999 and that drive too still functions and has 117 bad sectors on it per the last S.M.A.R.T. check on it. Can't speak for others, but that's my experience with IBM/Hitachi drives. I know that Western Digital purchased the HGST division a while back so now I have to wonder if I can continue to trust Hitachi drives in the future. This 640GB is an older model pre-WD acquisition.

As for the number of bad sectors per the hex raw value, it's hexadecimal, it's not some magical mystical thing, when you convert it to decimal that's what you have: the raw value and it's not subject to some funky interpretation. If it was, that would mean the standardized attributes for S.M.A.R.T. could be different for every manufacturer and if that were the case then what's the point of having the standardization in the first place? See how that works? :)
 
Tiberian said:
As for the number of bad sectors per the hex raw value, it's hexadecimal, it's not some magical mystical thing, when you convert it to decimal that's what you have: the raw value and it's not subject to some funky interpretation. If it was, that would mean the standardized attributes for S.M.A.R.T. could be different for every manufacturer and if that were the case then what's the point of having the standardization in the first place? See how that works? :)
Click to expand...

The raw values are device specific. Hence the name raw. Often they are simple counts, but not always. A generic SMART alarm bell program wouldn't use them, it would use the "normalized" value and threshold (higher is better).
 
Last edited:
Tiberian said:
You're fine, the drive is fine, but I've noticed you've been asking a lot of questions about hardware of late and so far whatever machine you bought or built appears to be working fine. Stop worrying so much and put the damned thing to use. :)
Click to expand...

I've been asking a lot of questions because I want to lean. Sometimes reading Wiki and other sources of information are overly complicated/need to be a damn IT pro to understand. And I'm considering taking the A+ and working either in sales or doing something else tech-related. Not only that, but I just bought 5x 10 TB HDDs which cost about $420 per, so I wanted to make sure they were ok before I tossed them in my NAS because I think the NAS only uses S.M.A.R.T data, so I've been scanning them via SeaTools, and Data Lifeguard in an external enclosure because I'm not sure if SeaTools/Data Lifeguard can be used via the NAS. The Ironwolf isn't going in the NAS though so no Iron Health, which seems to be a bit of a gimmick anyway.

And it might be kinda nice to build my own desktops/be able to fix my hardware/software issues from now on, or at the very least, know what tech I want and why I want it, beyond a very basic level of understanding. I kinda got off on the wrong foot with tech in my youth. First three PC's were DOA, and it was a pretty big turn-off from wanting to learn more - made me feel very timid/irritated around tech; trying to change that. Blah blah blah, that's my story. ;D
 
Tiberian said:
I
Click to expand...


i have not observed that on Hitachi drives that i have had in the past if i had and seen these items i would of replaced the drive right away just from the errors on 5 of them SMART fields (actual relocated events was Hex 7B witch is 123 on your picture which may also be incorrect but 123 sound more plausible than 7 million) in any case i would not continue to use a disk if relocated events have happened just asking for it to fail later on

on diskinfo i always change it to decimal as FenFox did (function > adv > raw values > Dec) or if using HDD sentinel change to decimal

FenFox said:
I've been asking a lot of questions because I want to lean. Sometimes reading Wiki and other sources of information are overly complicated/need to be a damn IT pro to understand. And I'm considering taking the A+ and working either in sales or doing something else tech-related. Not only that, but I just bought 5x 10 TB HDDs which cost about $420 per, so I wanted to make sure they were ok before I tossed them in my NAS because I think the NAS only uses S.M.A.R.T data, so I've been scanning them via SeaTools, and Data Lifeguard in an external enclosure because I'm not sure if SeaTools/Data Lifeguard can be used via the NAS. The Ironwolf isn't going in the NAS though so no Iron Health, which seems to be a bit of a gimmick anyway.

And it might be kinda nice to build my own desktops/be able to fix my hardware/software issues from now on, or at the very least, know what tech I want and why I want it, beyond a very basic level of understanding. I kinda got off on the wrong foot with tech in my youth. First three PC's were DOA, and it was a pretty big turn-off from wanting to learn more - made me feel very timid/irritated around tech; trying to change that. Blah blah blah, that's my story. ;D
Click to expand...

i normal disk should not be showing read error rate , reallocated sector count, relocated events , currant pending sector , uncorrected sector (01 05 c4 c5 c6) on that drive you shown (different drives will have slightly more or less info) some HDDs the 01 SMART "read error rate" is not an actual error and can have some large numbers in there as it norm same number is shown in corrected errors in another field

if you have no pending counts you could continue to use the disk as long as more sectors don't go bad (but personally i would not)

if your using synology NAS device don't think they use SMART much they use things like how long it has taken to read or write data if its taking longer for that to happen it mark a disk as bad

even if you test it externally and it passes there could be some really slow sectors that the drive is having issues with but the drive is internally managing to read the data evidently without marking a SMART event, HDD sentinel would likely show slow sectors on a Full read test
 
likeman said:
i would not continue to use a disk if relocated events have happened just asking for it to fail later on
Click to expand...

You seem to be missing the point: if anything on this Hitachi drive is defective it's the S.M.A.R.T. technology itself. :)

Hitachi's own DFT gives the drive a 100% clean status so, as far as I'm concerned that means the drive itself is functionally sound, I don't care what the S.M.A.R.T. status says about things because I have a history of drives from every manufacturer with clean S.M.A.R.T. status and yet the drives fail the manufacturer's diagnostics.

The only tool you can truly put any confidence or faith in for checking hard drives legitimately is the diagnostic for the drive provided by the manufacturer of the drive.

I've used SpinRite since practically the day it was created years ago and even with that fantastically simple and deadly efficient app (not as useful as it once was but still a good thing to have around in some situations) I don't trust it 100%, more like 98.5% compared to the manufacturer's diagnostic tool. MHDD is another great tool as well but that too isn't fully trusted like the manufacturer's tool is.

Anyway, that's my experience in a nutshell: I'll never ever trust S.M.A.R.T. for drive status aside from some parameters related to performance but not health aspects, if you want to put faith in it, go for it. :)
 
Tiberian said:
You seem to be missing the point: if anything on this Hitachi drive is defective it's the S.M.A.R.T. technology itself. :)

Hitachi's own DFT gives the drive a 100% clean status so, as far as I'm concerned that means the drive itself is functionally sound, I don't care what the S.M.A.R.T. status says about things because I have a history of drives from every manufacturer with clean S.M.A.R.T. status and yet the drives fail the manufacturer's diagnostics.

The only tool you can truly put any confidence or faith in for checking hard drives legitimately is the diagnostic for the drive provided by the manufacturer of the drive.

I've used SpinRite since practically the day it was created years ago and even with that fantastically simple and deadly efficient app (not as useful as it once was but still a good thing to have around in some situations) I don't trust it 100%, more like 98.5% compared to the manufacturer's diagnostic tool. MHDD is another great tool as well but that too isn't fully trusted like the manufacturer's tool is.

Anyway, that's my experience in a nutshell: I'll never ever trust S.M.A.R.T. for drive status aside from some parameters related to performance but not health aspects, if you want to put faith in it, go for it. :)
Click to expand...

you are correct when to say that smart says everything is ok when it's not so can't really trust SMART to say that your drive is good, but if your getting actual logged SMART events in the amounts of events logged that your getting on your drive i just would not use it (unless you dont care whats on it or is backed up) UltraDMA, command timeout and Relocation event (your Reallocated sector count seem broken or Hitachi's non standard way to log it)

i had disks failing that have not been logging any errors at all (all that the HDD tools use any way is DST short and long witch can be triggered from any program nothing specific special about it in manufacture software)

but if any SMART errors start showing on sector smart fields i just replace it, it should always be 0 if it's not your drive is developing problems that it is having to correct it self and should no longer be trusted
 
You must log in or register to reply here.
Back
Top