Hard drive reliability

[goontron]

Re: Hard drive reliability

Well, so far so good.... /s

[petehall347]

Re: Hard drive reliability

haven't read the thread but i think spinny disks will save data better than ssd . or does a head crash wreck the disks ? .

[momaka]

Re: Hard drive reliability

bad luck i guess. if the drive was sitting on the shelf instead of behind the counter, it may have gone through lots of fiddling and moving around with and someone else could have dropped the drive and put it back. and after a few days of power on, bam... PHAIL!

I think any physical shock when the drive is off can only cause a pass/fail scenario for the HDD right from the get-go - i.e. the hard drive wouldn't work as soon as you tried to power it on. Reason being, when the HDD is off and you give it excessive shock, that will only affect the heads. The platters are not moving and not touching anything (or rather, the heads are not flying over them), so the platters should be fine.

Ive tried zeroing it. It fails in the same spot as the RAID controller reported.

dd: writing to '/dev/sdh': Input/output error
1060737+0 records in
1060736+0 records out
543096832 bytes (543 MB, 518 MiB) copied, 13.3418 s, 40.7 MB/s

In that case, you can try running a HDTune quick scan and take screenshot of the results, if it finds any bad sectors. Then do a full scan and compare. Obviously if you do format around the bad sectors, make sure to use the results from the full scan. The only times I run a quick scan in HDTune is if the HDD drops out from the full scan. Managed to map-out a failing Toshiba 2.5 " HDD that way. It's fairly stable now.

[RJARRRPCGP]

Re: Hard drive reliability

Here's a real turd. First error is 543096832 bytes in...

Never buy the 5k RPM blue line of drives.

At sub-1GB! WTF! And dd only recorded a paltry-for-even-a-dozen-years-ago 40 MB/s!

Looks like a steaming pile o' dogshit!

[goontron]

Re: Hard drive reliability

Given the bad sectors so early in its life, most likely some parts of the magnetic media on the platters is marginal and probably needs to be mapped out. Similar to CPU and GPU binning, HDD platter binning probably works in a very similar way - i.e. if for example, an HDD of 1 TB with 2 platters and 4 heads (250 GB per platter side) has lots of defects on only one side of one of its platters, the manufacturer may just remove the head for tha platter and disabled it in firmware, effectively making the HDD a 750 GB one. If another platter has problems, they may even bump it down to 500 GB or even 250 GB.

So in essence, the platters in your HDD probably passed the factory tests enough to qualify as capacity X, but over time turned out to be unstable and the drive really should have been marked down as a lower capacity (or just junked altogether).

Thus, I only see this ending two ways: either the drive is going to reallocate some (or lots of?) sectors, then "settle down" and play nice -OR- it will reallocate until the whole drive is dead/useless and/or not detecting.

To get that process going, my suggestion is to perform a zero fill once. This should hopefully find all bad sectors that need finding, and then the HDD should either remap them or leave them as bad. After this, do a normal read/surface scan with HDTune and look where the bad sectors are. Take a screenshot and then re-run the scan. If the bad sector map stays the same, you can either format the HDD into partitions around the bad blocks (if they are fairly large and clumped together), or use the HDD as is, if they are all over the place, but not changing (if they are, the drive is useless). But if the bad sectors keep increasing every time or appearing in different areas from before, then no point in further trying to do anything with the HDD. (Actually, even this is probably too much for most people. But given it's a fairly big 4 TB drive, it might be worth a shot to keep around as an OS test drive. At least I would )

As I've mentioned before in this thread, a good number of my drives have at least a few bad sectors (some even several hundred!) But as long as the HDD settles down and the bad sectors stop increasing (or just do at an extremely slow rate), you can still get life out of that HDD.

The PC (laptop) on which I am typing this post from has a 20 GB IBM Travelstar with 5 bad sectors (though some of these might have resulted from a bad connection, as the drive also has 26 in the UltraDMA error count). I've been using it like that since 2010! And I'm not even its first owner - the HDD is probably from the very early 2000's. It's also 97% full at the moment - i.e. most of its sectors are in use with my data. Do I trust it? - After 8 years, I can safely say, yes - at least more than any new HDD from today.

Ive tried zeroing it. It fails in the same spot as the RAID controller reported.

dd: writing to '/dev/sdh': Input/output error
1060737+0 records in
1060736+0 records out
543096832 bytes (543 MB, 518 MiB) copied, 13.3418 s, 40.7 MB/s

Im going to run spinrite on it, on the off chance its just out of alignment....

[mariushm]

Re: Hard drive reliability

Modern hard drives seem to be more reliable.

Anyway... time to buy another 4 TB drive to replace the 1 TB drive....

The Seagate 2 TB drive died on me maybe 2-3 months ago... i expected it to die before my WD drive, still sad I didn't manage to backup some of the movies i had on it (nothing i can't download again but not easy)

The black WD 1 TB still is the champ with 66k+ hours:

[momaka]

Re: Hard drive reliability

Here's a real turd. Or maybe bad luck. I'm not sure. Its a product of what i like to call the WD "Turquoise" line. Where they merged the green line into the blue line of drives.

Model Family: Western Digital Blue
Device Model: WDC WD40EZRZ-00GXCB0
Firmware Version: 80.00A80
User Capacity: 4,000,787,030,016 bytes [4.00 TB]
Rotation Rate: 5400 rpm

3 Spin_Up_Time POS--K 209 186 021 - 4533
4 Start_Stop_Count -O--CK 100 100 000 - 14
5 Reallocated_Sector_Ct PO--CK 173 173 140 - 794
7 Seek_Error_Rate -OSR-K 127 013 000 - 3107
9 Power_On_Hours -O--CK 100 100 000 - 41
194 Temperature_Celsius -O---K 125 117 000 - 25
196 Reallocated_Event_Count -O--CK 145 145 000 - 55
197 Current_Pending_Sector -O--CK 200 200 000 - 12

It was online less than two days before my RAID controller kicked it offline for IO errors. First error is 543096832 bytes in...

Given the bad sectors so early in its life, most likely some parts of the magnetic media on the platters is marginal and probably needs to be mapped out. Similar to CPU and GPU binning, HDD platter binning probably works in a very similar way - i.e. if for example, an HDD of 1 TB with 2 platters and 4 heads (250 GB per platter side) has lots of defects on only one side of one of its platters, the manufacturer may just remove the head for tha platter and disabled it in firmware, effectively making the HDD a 750 GB one. If another platter has problems, they may even bump it down to 500 GB or even 250 GB.

So in essence, the platters in your HDD probably passed the factory tests enough to qualify as capacity X, but over time turned out to be unstable and the drive really should have been marked down as a lower capacity (or just junked altogether).

Thus, I only see this ending two ways: either the drive is going to reallocate some (or lots of?) sectors, then "settle down" and play nice -OR- it will reallocate until the whole drive is dead/useless and/or not detecting.

To get that process going, my suggestion is to perform a zero fill once. This should hopefully find all bad sectors that need finding, and then the HDD should either remap them or leave them as bad. After this, do a normal read/surface scan with HDTune and look where the bad sectors are. Take a screenshot and then re-run the scan. If the bad sector map stays the same, you can either format the HDD into partitions around the bad blocks (if they are fairly large and clumped together), or use the HDD as is, if they are all over the place, but not changing (if they are, the drive is useless). But if the bad sectors keep increasing every time or appearing in different areas from before, then no point in further trying to do anything with the HDD. (Actually, even this is probably too much for most people. But given it's a fairly big 4 TB drive, it might be worth a shot to keep around as an OS test drive. At least I would )

As I've mentioned before in this thread, a good number of my drives have at least a few bad sectors (some even several hundred!) But as long as the HDD settles down and the bad sectors stop increasing (or just do at an extremely slow rate), you can still get life out of that HDD.

The PC (laptop) on which I am typing this post from has a 20 GB IBM Travelstar with 5 bad sectors (though some of these might have resulted from a bad connection, as the drive also has 26 in the UltraDMA error count). I've been using it like that since 2010! And I'm not even its first owner - the HDD is probably from the very early 2000's. It's also 97% full at the moment - i.e. most of its sectors are in use with my data. Do I trust it? - After 8 years, I can safely say, yes - at least more than any new HDD from today.

[ChaosLegionnaire]

Re: Hard drive reliability

bad luck i guess. if the drive was sitting on the shelf instead of behind the counter, it may have gone through lots of fiddling and moving around with and someone else could have dropped the drive and put it back. and after a few days of power on, bam... PHAIL!

[goontron]

Re: Hard drive reliability

^ So have I. One of the drives in the array is a 2tb Green, 17699 POH, 0 sectors relocated or pending

I also picked up a 500Gb external Green drive from a junk shop in Colorado. 771 POH, 0 sectors relocated or pending. Belonged to a teacher at CSU. I actually contacted the previous owner and asked if they wanted any data from it before i wiped it. Unsurprisingly, since it was a backup drive, they turned me down. The proprietary PSU for the enclosure had died. It was a chineseum junk case, so that's not surprising.

[Topcat]

Re: Hard drive reliability

^
Strangely enough, I've had pretty fair luck with Greens....I've heard all the horror stories, but I've been using them in my NAS for years. Not so much as a single reallocated sector. OTOH, I went back to the parts room about a year ago and the RAID controller was screeching at me.....to find out one of the Seagate Cheetah SAS drives (ST3146356SS) had shit the bed. Replaced with a spare, rebuilt array, and life is good.

[goontron]

Re: Hard drive reliability

Here's a real turd. Or maybe bad luck. I'm not sure. Its a product of what i like to call the WD "Turquoise" line. Where they merged the green line into the blue line of drives.

Model Family: Western Digital Blue
Device Model: WDC WD40EZRZ-00GXCB0
Firmware Version: 80.00A80
User Capacity: 4,000,787,030,016 bytes [4.00 TB]
Rotation Rate: 5400 rpm

3 Spin_Up_Time POS--K 209 186 021 - 4533
4 Start_Stop_Count -O--CK 100 100 000 - 14
5 Reallocated_Sector_Ct PO--CK 173 173 140 - 794
7 Seek_Error_Rate -OSR-K 127 013 000 - 3107
9 Power_On_Hours -O--CK 100 100 000 - 41
194 Temperature_Celsius -O---K 125 117 000 - 25
196 Reallocated_Event_Count -O--CK 145 145 000 - 55
197 Current_Pending_Sector -O--CK 200 200 000 - 12

It was online less than two days before my RAID controller kicked it offline for IO errors. First error is 543096832 bytes in...

Never buy the 5k RPM blue line of drives.

The worst part about it, i was trying to salvage a failing array so the first thing i did was go down to Walmart and buy an external drive to pull the drive out of, instead of having one shipped, so no warranty for this thing....

[momaka]

Re: Hard drive reliability

Upload more!

Alright.
Ask, and you shall receive!

... well, I'm going to upload the sound for only one hard drive today. But, I figure the details I put into this post should make it more worthwhile, as I think this HDD should be quite interesting to listen / read about. Mr. Goontron may want to pay special attention here .

What is it?
- A Quantum Bigfoot CY 3240A (3.2 GB) 5.25" hard drive that I found many many years ago back in high school out of a discarded socket 7 Compaq PC (from which I still have the mobo and PSU).

While this hard drive still works, it's not without some issues. And it seems those issues were present since day one when I got it (more on that below, in the history of the HDD). As such, this HDD has not seen much use in my hands - maybe 10 hours total in the last 15 years, if even that much. But I will let the SMART log speak for itself:
https://www.badcaps.net/forum/attach...1&d=1521783361

For those who do not wish to login to view it, let me give you a summary: the HDD has only 223 total running hours up to date, but with 1018 power (On/Off) cycles. I'm not sure why this HDD has so many power cycles, but I can assure you they are not from me (I don't think I am responsible for even 1/3 of those).

The worst part is not the power cycles, though. It's the bad sector count, currently at over 183 (HDD found a few more after saving the SMART log last week). And moreover, I haven't even attempted to do a full sector scan with HDTune. I did only a quick one, and that revealed a bunch at the end of the drive. When I tried starting the full scan in HDTune, the HDD was finding bad sectors almost every 5-10 seconds, so I stopped the scan (it was late at night, and I didn't want to wait for it).

That said, I did get some very interesting sound recordings out of this Bigfoot HDD. Here is the first one after the HDD sat idle on my shelf without power for possibly more than 4-5 years:
https://www.badcaps.net/forum/attach...1&d=1521783361
When I heard the first part of this sound recording, I thought the HDD has sure finally gone to e-Heaven. For those confused what I am talking about or just plain curious what they are hearing in this clip, here's a timeline of the sound above:
@ 3 to about 9 seconds into the clip, HDD spins up to speed (3600 RPM, hence the loud 60 Hz "rumble").
@ approx. 10.5 to 11.5 seconds, the HDD makes a clicking noise, inditacting that headstack calibration is about to begin. But soon afterwards, a hollow metallic "bong" sound is heard with the heads retracting back into parking position due to a an error in calibration
@ approx. 13.5 to 14 seconds, the HDD attempts another calibration cycle. Here, you can hear the HDD head click several times before the hollow metallic "bong" sound is heard again, indicating that the heads have retracted and calibration failed once more.
@ between 16 and 17 seconds, as well as 17.5 and 18.5 seconds, the HDD attempts two more head calibration cycles like above. After the last failed one at approx. 18.5 seconds, the HDD gives up and spins down automatically.
@ approx. 29 seconds, I cycle power to the HDD to try another attempt at making it work.
@ approx. 36.3 to 36.75 seconds, there's another short failed head calibration attempt.
@ approx. 39.1 to 41.1 seconds, the HDD finally does a proper full head calibration and appears to work normally.
@ approx. 50.6, I disconnect power to the HDD, and it automatically parks its head (that same metallic "bong" sound) and spins down.

After this whole scene, I decided to record another boot up sound, thinking I might get a more clear recording. But as you can hear on the following recording, there are some weird stray bearing sounds - it seems that this HDD is also either starting to develop bad bearings or perhaps they were of poor quality to begin with.
https://www.badcaps.net/forum/attach...1&d=1521783361

With the this Bigfoot HDD finally able to spin-up without issues, I decided to connect it to a computer and run some benchmarks, as well as look at the SMART log (that I posted above). While the SMART log didn't surprise me much with the bad sectors, the HDTune benchmark did a little bit. Here it is:

... yup, 6.9 MB/s is the maximum transfer speed I attained with this drive. Sure makes me appreciate the speed of modern drives - even on not so modern ones, like the 80 GB Spinpoint that's in the computer I'm typing this from right now.

But as you can see from that graph, the HDD is indeed quite sick. During many times in the benchmark, this Bigfoot HDD made the same hollow metallic "bong" sound multiple times, indicating the HDD has hit a bad block/sector - and hence the 0.0 MB/s minimum transfer speed. You can also see the benchmark did not fully complete. This is because near 97-98% of the way, the HDD found too many bad sectors and thus making HDTune drop the benchmark. After this, I attempted two more benchmark tests, but they both failed much earlier.

After this, I ran a quick surface scan, and got more interesting sound recordings. Here you can hear how this Quantum Bigfoot HDD sounds when it finds a bad sector and attempts to reallocate/correct it:
https://www.badcaps.net/forum/attach...1&d=1521783361
The HDD finds the bad sectors around 8.5 to 9 seconds into the sound clip above and finishes correcting them at around 16 to 18 seconds. (You may also hear a slightly different background font. This is due to the HDD being installed in another computer - an HP Pavilion 8756c with a loud IBM Deskstar DTLA-305030 and rumbling CPU fan running at 5V for more "silence").

Now for a little history about the HDD:
I found it around 2004 or possibly early 2005, I believe. At that time, big flash drives (256-512 MB or more) were rather expensive, so I started using the Bigfoot above as my first "portable" HDD to move files between my main PC and another offline PC I had at my uncle's house. Before that, I used to burn CDs like crazy, as it was the only way I could move big files around (DVDs were out of the question, as DVD writers were still expensive at the time... at least to me). I used the Bigfoot a few times for about a year or so. Unfortunately, it didn't turn out to be too reliable of an HDD, despite its looks. In fact, I remember back when I first formatted it, the format took several tries and a very long time (almost 30 minutes, if not more). The HDD made strange noises on the first attempt as well. But after it formatted, it appeared to work normal, so that's why I continued to use it (NOTE: I was a complete PC noob at the time). Anyways, on the 3rd time I tried to use the HDD to move files to my offline PC, the HDD hung on a few files that I really needed. Since then, I retired the hard drive. I did still use it for a few test PCs a few years later, but not much at all by any means. By that time, I already had a 512 MB flash drive, as well as a 30 GB Quantum Fireball LCT 10 HDD that I scored on eBay for relatively cheap at the time.

In any case, that Quantum Bigfoot HDD was never reliable to me. Despite that, I elected to keep it simply because it looked and sounded too cool. So here it is many years later awoken to be heard once again. I hope you all enjoyed.

[itachi]

Re: Hard drive reliability

My main storage drive is a Western Digital Black 640GB which I got several years ago new. It has 70k hours on it with 2346 cycles. My second most used drive is another WD Black 1TB with 63k hours and 1705 cycles. Crystal disk info says they're both perfectly healthy.

I've worked on a lot of computers. I have a decent stack of failed toshiba, seagate and the odd samsung drive. Once I ordered a WB Green drive and failed with bad sectors while I was copying new data to it. I only get the black ones now as they seem to be the most robust.

[Pentium4]

Re: Hard drive reliability

Let me know if you'd like to hear some of my other HDDs. I have a 5.25" Quantum BigFoot, among a few other interesting drives.

And oh yeah... how do you like my ID3v1 tags on that mp3?

You should have seen the grin that appeared on my face after hearing that Netserver boot. That was so cool. Upload more!

Later I'm going to post more HDD pictures. I went through all of my failed drives and found a few that made it well over 70,000 hours.

[Dan81]

Re: Hard drive reliability

Ran a Crystal Disk Mark on my Samsung HD103SI.

Got a caution regarding Reallocated Sectors Count (100) but otherwise Power On Count is at 6782 and 11463 Power On Hours.

[RJARRRPCGP]

Re: Hard drive reliability

Wow! Some high POH hard drives in here. I have found the 7200.7 to be pretty reliable, as well as WD Black. Hitachi drives also hold up well.

I have a 160GB 7200.7 that I think is finally on its last legs. It's the one I posted on the first page. It has 66,583 POH and it's starting to act funny. It has 5 reallocated sectors, 115 pending sectors, and the seek error rate number gets higher after every 10 seconds when transferring big files. I was just going to let it run in my room but it will make strange noises periodically so I think it's ready to retire. It had a long life, but not as long as kevins!

I have a Hitachi IDE drive in my Pentium III pfSense machine and it has 68,600 POH. It has 0 reallocated sectors and 1 pending sector, but after 6,000 hours that number hasn't increased. It also only has 116 power cycles, and the seek error rate is 0. I'm hoping it will last another 10,000 at least. Haha.

Here is a poor Raptor drive. It has a good life but I couldn't do anything to repair it. It spins up fine but when you try and run a test on it, it immediately fails. I tried plugging it into Windows and it won't show up in Disk Management. I was mostly bummed because I wanted to see the SMART statistics. The only thing I was able to see was the POH, which is a decently high number.


This drive when you plug it in won't make any noise but the logic board gets scalding hot really quickly. Couldn't figure out any stats on it.


This drive worked the first time I tested it, with around 23,000 hours if I remember correctly. It got stuck on a bad sector once and then the next time I tried to look at the data, it would spin up, then spin right down. Couldn't hear any arm movement.


These are the highest POH drives I've ever seen. They ran side by side in a Windows 2000 Advanced Server, well into 2015. They never failed! Although one of them has 8 reallocated sectors. Not too bad for that much use!


I have 80 drives I have tested so far and these are the only ones that have failed. I have a few more, but so far that's pretty good.

I suspect a bad board or component on the board of the 2007 Raptor...
Especially if Windows don't show anything or all tests immediately fail...

[momaka]

Re: Hard drive reliability

^ DO IT!
The other day, I was thinking to record the sounds of some of my PCs as well. Obviously, my new systems are pretty boring and I probably would skip all of those. But I really need to make a better recording of *my* (own) first PC (a socket 462 with 2 old HDDs and 2 ODDs). Stay tuned, I suppose.

[goontron]

Re: Hard drive reliability

Okay, I found a cool sound file recording of my movie/games/files backup rig #2, and I couldn't resist posting it. It's my HP NetServer E800 - basically a stock NetServer E800 tower with a 80 GB Maxtor Atlas IV (ATLASU320_73_WLS) and a 300 GB Seagate Cheetah 10k7 (ST3300007LC) for the HDDs, a Teac CD-540E CD-ROM, and also a Toshiba(?) DVD-RW drive.

More info about what's going on in the sound file:

* The first 4 clicks you hear in the video are of me pressing the power button. I'm not sure if it's the PSU or the motherboard, but the PC won't always start on first power button press (the fans kick in for a second and then stop, like a shorted/overloaded power supply).

* after this (@ 10 seconds into the video), the Teac CD-540E CD-ROM drive cycles its tray and seeks for CD. 2 seconds later, the DVD-RW drive also does a disc-seek.

* @ approx. 16 seconds, the Maxtor Atlas IV begins to spins up.

* @ almost 20 seconds, the Maxtor Atlas IV is pretty much up to speed, so you hear a *click* that enables the heads to move.

* @ 29 seconds, Maxtor Atlas IV completes its spin-up + head-seek sequence (what a lovely noise! )

* @ approx. 35 seconds, Seagate Cheetah 10k7 begins to spin-up (I have the SCSI HDDs on staggered spin-up, since my PSU is only rated for 6 Amps on the 12V rail)

* @ approx. 54 seconds, Seagate Cheetah 10k7 has completed its spin-up and head-seek sequence

* @ approx. 59 seconds, a former neighbor slams his car's door

Let me know if you'd like to hear some of my other HDDs. I have a 5.25" Quantum BigFoot, among a few other interesting drives.

And oh yeah... how do you like my ID3v1 tags on that mp3?

I should post my systems coming up after a power drop. Bigfoots, 10k and 15k SAS, fans... Nice sounds.

[TechWizard.Support]

Re: Hard drive reliability

I've had a few of the WD Blue 3200BEVT drives fail recently with less than 10,000 power on hours! The older drives do tend to last much longer it was common to see 50,000+ hours before disks failed. I think it's to do with the higher data density as previously mentioned.

[goodpsusearch]

Re: Hard drive reliability

All this sequence of sounds makes you feel you have some serious machine and not a toy, as with today's fanless laptops with SSD.