Hard drive reliability

[eccerr0r]

Re: Hard drive reliability

Ah forgot that some drives actually printed the number of sectors on the label... doh!
Yeah it's a FYPS so that's why it's slow. Alas it's got too many bad sectors to be a 2TB drive anymore.... now using about 1.7TB which so far, so good, no more failures for now. Using it as a rsync backup disk for ... my WD Green drive ...

Was using my RAID array for backup for the WD Green drive, but now I can use it for something else...

I wonder if I should try to set up another RAID5 with this and two other 2TB disks... Basically RAID5 the known good sections of the disk, and make a RAID1 of the two good disks where the third disk is known bad... Maximizes the use of the good sections of all disks?

[momaka]

Re: Hard drive reliability

Here's the highest hours I've seen so far. This hard drive is badass. It had 90,000 hours when I got it. It was in a, wait for it.... SFF case with a Pentium D! Ouch! And the fan PSU failed, but it kept chugging along with bulged caps for who knows how long. So I put it in my IPFire box even though it had reallocated sectors. From the 90k point to this picture (June 18, 2021), it only gained 6 more reallocated sectors.

Holy-moly! 119k POH?!

I always knew those 80 GB Seagate Barracuda 7200.7 HDDs were built like tanks... but that lifetime really is something you don't see every day.

Strangely enough, I only find the 80 GB ones to be that reliable. The 120 GB 7200.7 seem to not make it as far (probably because they have an extra platter and head to deal with... and perhaps as things get "loose" inside over time from wear, the headstack is more likely to suffer from misalignment failures.)

As for the few bad sectors - that's actually fairly "normal" for those 7200.7 HDDs. Same with the Barracuda ATA III and IV series - some will never get any bad sectors, but other might add a bad sector or two once in a while. I suspect as the magnetic surface of the platters ages, the drives with less-than-perfect coating is the reason why we see bad sectors on some of these. But at least from what I've noticed, the rate of the bad sectors increasing on these HDDs tends to be so slow that they could probably easily run for another 10 years, if the other parts (headstack assy. bearings, motor bearings) can make it that far. For an HDD running 24/7 (but not doing much R/W), I would guesstimate the motor bearings would wear first and cause alignment issues for the headstack due to excess vibration or movement. And for an HDD doing lots of R/W, the headstack bearings might fail first. As for start/stop cycles and/or power cycles - I think those also do some wear on the heads themselves when the heads go to the parking zone on the platters... but all in all, most of these 7200.7 drives can easily sustain 3000-5000 power-cycle counts. I tend to be a hard "power-cycler", since I often turn off my PCs when they're not in use. Thus, a lot of my HDDs gain approx. anywhere from 300 to 1000 power cycles per year (laptop drives even more.)

That P4 box moved over ~63TB of network traffic while in service.

Dang!
In that case, keep it in the closet. If the Internet on this world ever goes down for any reason (WW3, next-gen Covid pandemic, etc.), I think we can easily count on such old boxes to "temporarily" keep things going.

I swear, those Pentium 4 CPUs are indestructible! Intel may have goofed up on their design, in terms of architecture and how inefficient they are for most tasks. But they sure have more than made up for it with their robustness.

I know it I have so many PII/PIII/P4 S478 that I don't know what to do with.

Good!
Then don't do anything with them... yet.
Another 10-20 years, the retroPC community will probably appreciate these even more. And best of all, since they tend to be built like tanks, they will probably still work fine. So consider them a retirement "stock".
... and of course, if you do want to play around with them in the mean time as retro PCs... you still can, as that probably won't degrade from that (if anything, it would be better for the caps.)

Also, the board pictured above has an AGP slot. Initially, mistook it for one of those Dell systems that lack an AGP slot... which makes them not-as-valuable for retro builds. But since yours does, that would make an awesome Windows 98 build for 90's games. P4 @ 2.0 GHz will handle virtually anything from that era with 60+ FPS.

i just use them till they drop. thats the point of the hardware. all hard drives will fail eventually... just a question of not if but when.

^ This.

wonder if i should chance used 55k or 72k POH used hard drives...and how much should be paid for them...

Depends on the drives.
Server HDDs tend to be built for a higher standard usually... but over 50k hours for a 10k or 15k RPM HDD is going to have a lot of wear on the motor bearings. If it's a 7200 RPM, probably still has plenty of useful life.

As cheap as large drives are these days, it's hard to beat simple raid1 arrays. Most reliable and easiest to recover in the event of a failure.

True.
But that's also assuming you have (or can get) another system (PC) just like it, in the event the HDDs are OK, but something happens to the system.

The one thing I absolutely hate about RAID is that you can't (easily... or at all) take a RAID config from one machine and move it to a totally different machine and expect things to work.

So for this reason, I prefer just simple, common interfaces like SATA or PATA, so that I can take my HDDs to any other system and pull the data off of them there without any proprietary hardware required. On the other hand, this is more work, since if not doing a RAID setup, then one has to manually backup files (or rely on backup software) regularly to prevent data loss. In my case, though, I have most of my important data already backed up. So at worst, I might loose some bookmarks and maybe recent datasheets. Most everything else I keep a backup of on more than one HDD/drive/device.

Usually the drives with higher data density die much sooner than the old drives.

Indeed.
Though some really really old (<5 GB and smaller) HDDs had a lot of reliability issues. Like the Pentium 4 machines shown on the previous page, I think HDDs also had a "golden" era (or a few) in a similar way, in terms of reliability... and I think those 7200.7 Seagates were right-smack in the middle of it... or one of the more notable ones.

Hmm...also notice that it's losing a lot of sectors about a bit past 75% of the disk... I wonder if I should just tell the drive to behave like a 1.5TB disk or partition away the bad spot... and wonder how much longer I can use the drive with it like this...

I've done this and surprisingly got (and still do) a lot of life out of several such HDDs.

If the bad sectors are in the middle of the drive (or several clusters with good clusters in between), you create the partitions in order of encounter for both the good partitions and the bad ones. Then just give the bad ones a name that you can remember, so as not to use them (e.g.: "Bad-1", "Bad-2", and etc.) As long as you avoid writing files to those partitions, the HDD should behave itself most of the time. Also give the bad partitions a little bit of a "buffer" before their start and end, in case they do develop more bad sectors.

[eccerr0r]

Re: Hard drive reliability

I've been making partitions and then merging the partitions back together into one large one through volume management. Alas the growth rate of new bad sectors is unclear, seems I may need to cordon off a large portion of the disk, perhaps all blocks after the middle of the disk, which is unfortunate as half the disk would be unusable...

[Pentium4]

Re: Hard drive reliability

How the hell is this thing still alive???

It's a 2.5" drive made in August 2012. I'm currently beating it up with torrents, so I don't believe it'll last much longer lol.

[Pentium4]

Re: Hard drive reliability

how many hours before people typically give hard drives a retirement party?
wonder if i should chance used 55k or 72k POH used hard drives...and how much should be paid for them...

As long as the SMART data is healthy, I leave them alone. Since a mechanical failure can technically happen to any HDD at any moment, I don't let it bother me. As long as it's backed up, I don't mind even running important data on a high POH disk. I have some drives with over 50,000 POH running critical stuff. As long as they stay cool and back up, I don't worry. I like to use things to their full life. I work from home so I can usually get things back up and running quick in case something happens.

I cycle out drives with critical data usually every 5 years....and to further reduce data loss risk; also run them in RAID arrays for redundancy....but you're right, nothing lasts forever....which is why I can't preach enough for people to keep backups on more than one media; stored off-site.

Yes this is a good method, especially if you don't have the time to deal with a catastrophic failure.

Yay? Been running badblocks -w on this disk until it stops puking bad blocks. It hasn't stopped yet.

 5 Reallocated_Sector_Ct  0x0033  169  169  140  Pre-fail Always    -
    241
 7 Seek_Error_Rate     0x002e  200  200  000  Old_age  Always    -
    0
 9 Power_On_Hours     0x0032  040  040  000  Old_age  Always    -
    44419
196 Reallocated_Event_Count 0x0032  001  001  000  Old_age  Always    -    214
197 Current_Pending_Sector 0x0032  200  200  000  Old_age  Always    -    0

Was initially trying to reuse this disk in my software RAID5 but this is too unreliable now. Now it's been taken out of the array and replaced with another old disk.
I'm hoping I could squeeze more life from this disk and use this as a backup disk if it can somewhat still reliably retrieve information, or at least if it won't fail at the same spot as the other disk being backed up. Then again if one completely fails it's a problem (which would be implicitly a failure at the same spot...or technically, all spots.)

It's really sad but I would give up on that disk. Triple digit numbers on those critical ones are always risky. I've even seen drives under 50 reallocated sectors sit there for a long time. 100+ usually means more than a basic defect. In my experience, of course.

Check this out:

Hitachi HDS721050CLA362 (Manufactured December 2010)

ID                Current Worst  ThresholdData    Status  
(01) Raw Read Error Rate     100   95    16    0     Ok    
(02) Throughput Performance   135   100   54    99     Ok    
(03) Spin Up Time        118   100   24    12648642  Ok    
(04) Start/Stop Count      100   100   0    742    Ok    
(05) Reallocated Sector Count  100   100   5    131074   Ok    
(07) Seek Error Rate       100   100   67    0     Ok    
(08) Seek Time Performance    138   100   20    31     Ok    
(09) Power On Hours Count    93    93    0    49850   Ok    
(0A) Spin Retry Count      100   100   60    0     Ok    
(0C) Power Cycle Count      100   100   0    742    Ok    
(B7) (unknown attribute)     100   100   0    0     Ok    
(B8) (unknown attribute)     100   100   97    0     Ok    
(B9) (unknown attribute)     100   100   0    65535   Ok    
(BB) (unknown attribute)     94    94    0    6     Ok    
(BC) (unknown attribute)     100   98    0    40370990  Ok    
(BD) (unknown attribute)     100   100   0    0     Ok    
(BE) Airflow Temperature     74    58    0    404619290 Ok    
(C0) Power Off Retract Count   100   100   0    845    Ok    
(C1) Load Cycle Count      100   100   0    845    Ok    
(C2) Temperature         230   166   0    524314   Ok    
(C4) Reallocated Event Count   100   100   0    2     Ok    
(C5) Current Pending Sector   100   100   0    0     Ok    
(C6) Offline Uncorrectable    100   100   0    0     Ok    
(C7) Ultra DMA CRC Error Count  200   199   0    1     Ok    

Power On Time     : 49850
Health Status     : Ok

This drive was installed at 39,000 hours. I used it as a scratch disk. It got used hard in the 10,000 hours I had it when running. I moved about 16TB to and from it in that amount of time. It was perfectly fine and I noticed suddenly a huge drop in transfer speed. It dropped to like 5MB/s. I looked to see ~3000 reallocated sectors so I immediately started pulling everything off of it. The data above is right after all the data was transferred!

[momaka]

Re: Hard drive reliability

How the hell is this thing still alive???

It's a 2.5" drive made in August 2012. I'm currently beating it up with torrents, so I don't believe it'll last much longer lol.

https://www.badcaps.net/forum/attach...1&d=1645075632

Maybe it's counting the minutes and not hours?
In that case, 2 389 585 comes out to about 39.8k hours.
I have a few old Hitachi/IBM and Maxtors that count in minutes and not hours.
Just keep HDTune open and see if the "POH" increments by "1" every minute or so. If yes, you have an HDD that counts "POH" in minutes. As for why or who thought this was a good idea... I haven't a clue.

This drive was installed at 39,000 hours. I used it as a scratch disk. It got used hard in the 10,000 hours I had it when running. I moved about 16TB to and from it in that amount of time. It was perfectly fine and I noticed suddenly a huge drop in transfer speed. It dropped to like 5MB/s. I looked to see ~3000 reallocated sectors so I immediately started pulling everything off of it. The data above is right after all the data was transferred!

Well, glad you were able to catch it in time.
Looks like it lived through its useful life. Moving 16TB worth of data is worthy for retirement. My guess is the ball beating(s) in the headstack assembly have started to fail from so much moving back and forth. The ones that came from this Seagate ST31500341AS (1.5 TB 7200.11 series) had what appears to be very rough bearings on the headstack. It was used for video surveillance recording... so constant writing. But no worries! Now it's turned into a bench grinder, where the headstack is just not needed anymore.

Also, I thought I had posted this here... but apparently not. Many years ago, I used a 40 GB 2.5" Seagate laptop HDD in one of my desktops. It came from either an Xbox 360 or PS3 (not sure which) and already had over 1700 bad sectors. I used it for about a year or so before the PC I used it in went unstable (and it wasn't the HDD causing the instability, believe it or not.) SMART info and other stuff for that HDD can be found in this post:https://www.badcaps.net/forum/showpo...0&postcount=70
... along with how I managed to temporarily "fix" a Fujitsu HDD with a stuck head on the platter to the point where I could actually pull some data off of it.

[eccerr0r]

Re: Hard drive reliability

Yeah this 2TB disk probably had it. Seems most sectors past the center of the disk are unreliable. sector count keeps going up.

 5 Reallocated_Sector_Ct  0x0033  149  149  140  Pre-fail Always    -    408
 9 Power_On_Hours     0x0032  038  038  000  Old_age  Always    -    45537
196 Reallocated_Event_Count 0x0032  001  001  000  Old_age  Always    -    364
197 Current_Pending_Sector 0x0032  200  200  000  Old_age  Always    -    149

Sigh...2TB disk. Could use the space but risky -- very risky...

[Pentium4]

Re: Hard drive reliability

Maybe it's counting the minutes and not hours?
In that case, 2 389 585 comes out to about 39.8k hours.
I have a few old Hitachi/IBM and Maxtors that count in minutes and not hours.
Just keep HDTune open and see if the "POH" increments by "1" every minute or so. If yes, you have an HDD that counts "POH" in minutes. As for why or who thought this was a good idea... I haven't a clue.

The POH counter there is accurate, I watched it I was referring to the Load Cycle Count. Aren't drives usually rated for 300-600k? 2.3 million load cycles is insane, I'm just surprised that the head hasn't worn out. I hate power saving features. It hasn't gone up one since changing the power settings.

Well, glad you were able to catch it in time.
Looks like it lived through its useful life. Moving 16TB worth of data is worthy for retirement. My guess is the ball beating(s) in the headstack assembly have started to fail from so much moving back and forth.

Yes this is after it ran those 39,000 hours at a point of sales system at a car dealership open 7 days a week, so it got some heavy use.

The ones that came from this Seagate ST31500341AS (1.5 TB 7200.11 series) had what appears to be very rough bearings on the headstack. It was used for video surveillance recording... so constant writing. But no worries! Now it's turned into a bench grinder, where the headstack is just not needed anymore.

Wow!!! That's insane That poor thing haha... nice job on the grinder too, I love it!!

Yeah this 2TB disk probably had it. Seems most sectors past the center of the disk are unreliable. sector count keeps going up.

 5 Reallocated_Sector_Ct  0x0033  149  149  140  Pre-fail Always    -    408
 9 Power_On_Hours     0x0032  038  038  000  Old_age  Always    -    45537
196 Reallocated_Event_Count 0x0032  001  001  000  Old_age  Always    -    364
197 Current_Pending_Sector 0x0032  200  200  000  Old_age  Always    -    149

Sigh...2TB disk. Could use the space but risky -- very risky...

Yeah, that's super unfortunate. RIP! Does this happen to you often or do you just have a lot of drives?

[Pentium4]

Re: Hard drive reliability

This is the most POH I've accumulated on a SATA drive. This was a few years ago. I believe it was around 109,000 last time I checked. In perfect health! It's been with the same Pentium 4 the whole time. What an amazing system.

[RJARRRPCGP]

Re: Hard drive reliability

May be a problem with the board. 5 MB/s means possibly in PIO mode, which usually doesn't happen with SATA. Do you have any ATA communication errors?

[eccerr0r]

Re: Hard drive reliability

Yeah, that's super unfortunate. RIP! Does this happen to you often or do you just have a lot of drives?

Though I have a bunch of drives, no this is that same one drive I've been talking about, the FYPS WD enterprise drive. It was the replacement for another 2TB Toshiba drive that also went flaky but it seems to be still usable - perhaps it has something to do with the hotswap slot. Then again it's still correlation not causation - the FYPS drive was flaky well before it being placed in that slot where the 2T Toshiba went flaky. Using that Toshiba in another array at the moment where it seems to be doing okay unlike this FYPS.

I think I have 7 2T disks at the moment, one pretty much croaked (the FYPS), one that went screwey (the one Toshiba), three Toshibas that are okay, a Seagate that's okay, and a WD Green that seems to have exactly one pending bad sector at the moment. I haven't been able to find that bad sector yet...

[kevin!]

Re: Hard drive reliability

This is the most POH I've accumulated on a SATA drive. This was a few years ago. I believe it was around 109,000 last time I checked. In perfect health! It's been with the same Pentium 4 the whole time. What an amazing system.

The 9/10 series came out very well, and I'm sure it can do at least 50,000 POH more hours.
Another series that is doing well is the 1BD142, they are very reliable too.
Mine goes for 77.991h POH.
Although I don't know if the CPU will be able to withstand so much use due to the degradation of the silicon.

[eccerr0r]

Re: Hard drive reliability

I no longer have any < 500G disks that are running full time, too small now
Only have 2TB's powered up and spinning 24/7. Have a spun down 750G, however.

Technically POH keeps going up even if you have the disk spun down, I wonder if I can get my POH up higher with a disk spun down and never used...

[Heihachi_73]

Re: Hard drive reliability

8A4A = 35402. A long way for this one to get to the 140K hours of my 40GB Seagate (and my dead 60GB which was about 20K lower). Also, XP!!!!

[SluggerB]

Re: Hard drive reliability

I'm typing this from a system with a WD Caviar Black HDD that has t be over 15 years old now. I've always had excellent luck with platter HDD's

[dmill89]

Re: Hard drive reliability

Pulled this out of an old ThinkPad R61, a circa 2009 250GB Western Digitial Scorpio Black with 49,355 power on hours, 1,325 power cycles (I'm guessing this was the standard corporate "keep your laptop powered on after hours to receive updates" scenario, we do that at work too, but we don't keep our laptops long enough to hit these kinds of hours and have ordered laptops with SSDs since 2014). It has 18 reallocated sectors, so not in perfect health, but still functional and not making any "strange" noises.







And a 250 GB Seagate Barracuda 7200.12 with 48,192 power on hours and only 68 power cycles (definitely a corporate "leave your pc on for updates" situation) pulled out of a Dell Optiplex 390. This one does appear to be in good health and isn't making any "weird" noises either:

[nobbnobb1]

Re: Hard drive reliability

Here's something unique. Back in the day (pre-2010) I got a bunch of these Seagate EE25 Extreme Environment industrial grade drives for cheap. They are rated for 24/7 use and under vibration.

I built a Windows Home Server back then and at some point in time put one of these drives in as the OS drive since I figure why not use an ultra reliable drive for that. It's running Windows 10 now and is super slow, but for a server boot drive it's fine. Only 47K hours on it...the OS will likely become out of date before this drive fails.

[Dan81]

Re: Hard drive reliability

Some of the 2010 and the 2012 Seagate drives (the original 7200 series, to be exact) seem to be really reliable in my opinion. I have one 7200.10, two 7200.11 and two 7200.12, all ranging from 2007 to 2012. None of them has gone awry and I suppose these will still live to another 10-15 years.

[eccerr0r]

Re: Hard drive reliability

Thought that starting with the 7200.10 series, Seagate started to add planned obsolescence that they knew in advance that it'd run out of some lubricant or something after a while. Don't recall the details but my 7200.10 disks seem fine as of now but who knows ...

[Dan81]

Re: Hard drive reliability

Mine has about one relocated sector from what I remember. ST3500630AS, dated 1st of Feburary 2008, according to this site.

I'm thinking of doing a late 2000s (08' or 09') build with it as I love the sounds of this drive, and especially since I just got a free Plextor PX-800A (which is, as sad as it sounds, a rebranded Sony/NEC AD-7170A) that I'd also like to clean and use.