current status on HDDs?

[Gianni]

Re: current status on HDDs?

@ jpdoe: thanks for explanation. HDD Health is not from WD, it is from panterasoft.
Smart is smaller and simpler than HDD Health and run from DOS.
I have seen they give different values on the same item so I don't know which is correct.
The official soft from WD is Data Lifeguard but I can't see any information on HDD status.

@ i4004: thanks for the long doc, I will try to read it completely just to satisfy my curiosity.

@ dood: yes back up is the only thing to do, no just only when you doubt about HDD, but frequently

Ciao
Gianni

[zandrax]

Re: current status on HDDs?

@ Gianni: trust Smart raw values. According to them, you turned the drive on 1783 times but started and stopped 2591 times and standby-ed 2016 times (standby turn off the motors but not the electronics while other low consumption states keep the drive somewhat active, so they count as a start/stop event: confusing, isn't it?). It has been on for 3404 hours and 14 correctable error were corrected.
According to smart, the drive has nothing wrong: it can last up to 50000 start and stop cycles so it's far from being worn. Main concern is CRC affair: it may be caused by a single power failure or a damaged cable or, worse, being the first step for a slow platters corruption (first there are correctable errors, then they grow too much and the drive start developing uncorrectable errors which turns into bad sectors). I'm for the accident option because ECC correctable errors is empty; anyway check the cables and run HD Tach: the graph should be a parabole with an horizontal axis of simmetry and very little if no at all spikes. If there are some big spikes which slows down the transfert rate noticeably, then you're facing the early signs of corruption.

Zandrax

[kc8adu]

Re: current status on HDDs?

i see enough bad news in those screenshots to back that drive up now and retire it.

[jpdoe]

Re: current status on HDDs?

Hi, Gianni.
I took a quick look into western digital support web page. It seems there's two utilities: Data lifeguard tools (for assisting on installing the drive) and Data lifeguard diagnostics. The latter one seems to be the one with the SMART monitoring functionality.

[Gianni]

Re: current status on HDDs?

@ zandrax: I will try HD Tach one of these days... Thank you for suggestions.
I have a lot of things to do and no time as usual. Damn I have to get up within 4 hours and I have still something to do before going to bed...

@ jpdoe: Thanks!!! I'm getting older and probably I need glasses otherwise I don't see interesting things even if those are in front of me.
I'm downloading the diagnostic tool and ASAP I will give it a run.

Good night people
Gianni

[jpdoe]

Re: current status on HDDs?

You're welcome, Gianni. I took a look because it has happened to me too. HDD manufacturers put different functionality on different programs on a whim. IIRC, Seagate did the same thing.

[Gianni]

Re: current status on HDDs?

I have installed the diagnostic tool and it says my HD is OK.

Well I believe it... but just to be serene I use True Image from time to time

Ciao
Gianni

[i4004]

Re: current status on HDDs?

gianni, do you have external(usb) wd drive?
would this diag. program work with those?

recent versions of everest do(show smart data on them), but i dunno would this...

[Gianni]

Re: current status on HDDs?

Hi i4004

no I don't have an external WD HD and this is the only WD I have.
I have 3 Maxtor and one Conner .
I have also a small USB HD from LAcie but I don't know what it has inside, tomorrow I look for it and I will check.

Ciao
Gianni

[Gianni]

Re: current status on HDDs?

@ zandrax: here the test with HD Tach, as you can see there are 3 big spikes.

anyway check the cables

I forgot to mention I use mobile tray and I don't know if they can compromise HD life or test results.

@ i4004: you are lucky my external HD is WD and the diagnostic program works on it, look at the pictures attached.

Ciao
Gianni

[i4004]

Re: current status on HDDs?

>If there are some big spikes which slows down the transfert rate noticeably, then you're facing the early signs of corruption.

or just heavy fragmentation?

[Gianni]

Re: current status on HDDs?

No, the disk is de-fragmented so i think it is not the case.

Ciao
Gianni

[kikkoman]

Re: current status on HDDs?

fragmentation doesn't have anything to do with these surface read speeds.

there's another nice program called HDDScan that can even show detailed access times (for each block).



there's a plethora of hdd tools and most of them do the same thing. there's one tool (can't remember the name atm) that monitors when a SMART attribute changes and tries to predict when the next change will occur /when the drive will fail.

hard to tell if it's worth all the trouble. i usually have only speedfan running, that's all.

[jpdoe]

Re: current status on HDDs?

Actually, on these tests there's something similar to fragmentation.
When an HDD has reallocated sectors, the data thoroughput on a secuential read will take a dive when those sectors have to be read. That's because the heads will have to move to the spare sector used to reallocate the bad sector.

[i4004]

Re: current status on HDDs?

gianni, thanks!

fragmentation doesn't have anything to do with these surface read speeds.

news to me.
thanks

When an HDD has reallocated sectors

not a good news for gianni, as it means hdd is reallocating sectors...
but hold on...his smart says 200 for realocated sector count, while threshold is 140, and while "worst" is 200 too.
looking at those numbers i would say smart is not reporting correctly.

here's my smart report from everest(can find it on avax.ru site...beware, it can be hazardous site..go with secure browser and antivirus software) and hdtach...
"podaci" means "data" and by the looks of it it's accumulated data since drive was made, or so...

[i4004]

Re: current status on HDDs?

2nd seagate 7200.7 (this one is 120gb)
hooked to via pci ata-controller card

[dan8139]

Re: current status on HDDs?

Hi,

I do lots of stuff with drives and would like to say that running benchmarks / tests under Windows is great until something else accesses the drive...

MHDD is really good at showing disk surface condition.

What I do before reusing/reselling a drive:

Full scan with MHDD, if there are slow (>100ms) sectors then be on guard.

Full zero erase with Disktool, then if slow/bad sectors reported before a rescan with MHDD.

Then I look at SMART info... Reallocated Sector Count is the main one.

Sometimes lots of sectors just get 'slow' and the drive's firmware won't reallocate them, which is annoying!

I've got quite a pile of dead / dying HDDs, mostly Maxtor! If I get a system in with a Maxtor in (not only the slimline ones, Diamondmax Plus 9 also), I can pretty much guarantee it'll fail the MHDD test. Most of them seem dead / dying by now.

Perfect HDDs in use around here are Hitachi Deskstars, WD Caviar SE, Hitachi Travelstars. None are that old.

[jpdoe]

Re: current status on HDDs?

@i4004:
Regarding Gianni screenshots,
"worst" means lowest value the parameter took lately.
"threshold" is the lowest value the parameter can take without causing the smart test to fail.
"value" is the parameter value normalized to the range 0-253, right now.

So it's normal that "worst" is higher than "threshold".

The most interesting column is missing. smartmontools monitoring package shows it under the label "raw value", IIRC. It tells you the parameter value without being normalized to the range 0-253. For some rows it has an actual physical meaning; for "reallocated sector count" it tells the actual number of sectors reallocated so far.

I don't know what the "warranty" column means.

A few reallocated sectors is nothing to worry about.

[i4004]

Re: current status on HDDs?

>Regarding Gianni screenshots,
"worst" means lowest value the parameter took lately.
"threshold" is the lowest value the parameter can take without causing the smart test to fail.
"value" is the parameter value normalized to the range 0-253, right now.
So it's normal that "worst" is higher than "threshold".

that makes no sense to me. first of all how can "worst" mean lowest values?
aren't we counting errors, so that higher value means worse result?

2nd, threshlod should be highest value parameter can take without casing damage.
did you just swap word "highest" for "lowest" there?

then the value: but how can it be exactly 200?
on my drives that value is 100 on both drives?
so smart is telling me both drives have same number of realloc. sectors?
come on....
both drrives have 0 bad sectors.
also, for the 2nd drive i posted the results threshold is 0 for all fields...
and that makes sense as much as any other smart field..hehe...

i mean, this can't be read(aside the data value)..it just seems like random numbers that have no connection to reality.

>The most interesting column is missing. smartmontools monitoring package shows it under the label "raw value", IIRC. It tells you the parameter value without being normalized to the range 0-253. For some rows it has an actual physical meaning; for "reallocated sector count" it tells the actual number of sectors reallocated so far.

still makes no sense: gianni's drive has 0 reallocated sectors (raw section)
https://www.badcaps.net/forum/attach...achmentid=8894
and yet we see spikes in hd tach, and those can only be explained by reallocated sectors, right?
as we concluded fragmentation makes no difference...

i'm just looking at "value" and "worst" for power on time count on my hdds...
isn't that just silly?

i'm looking at spinup time....value is 98 and worst is 97.
no dummy, worst should be 98 then too!
update yourself already, you worthless piece of crap!
no wonder google couldn't find relation between smart and reality...
hehe...

and bonez said smart is smart.
quite...

overall the only usefull fields seems "data"(raw).

[jpdoe]

Re: current status on HDDs?

Let's suppose the HDD has 400 spare sectors.
How could the HDD designers have programmed the SMART firmware to calculate the other parameters?
They could have done the following:

Let's say they decided on a threshold value of 53. That will be the actual normalized parameter value when the raw parameter value is 400.

When the raw parameter value is cero (no spare sector used, still no reallocation events happened) the actual normalized parameter value will be 253.

253-53 = 200

So the firmware has a range of 200 to account for the hard disk tear and wear regarding reallocated sectors.

------------------------
The "worst" value can be thought as a memory that lets you see how low the normalized parameter value went in the past. It's useful for parameters that can go up and down (example: working temperature)

PS: I call "normalization" the act of translating the raw value (actual number of sectors reallocated) to the range between the parameter threshold (let's say 53) and 253.