My experience is that Seagate drives tend to be less than reliable. I've only had one and a half failures with WD in the past 14 years across 20 drives.

Friend of mine bought a 3TB Barracuda 14 months ago, and it flat-out died at 11 months. RMA'ed it with Seagate and got a replacement, and now it hasn't even been 3 months and it is already starting to fail again.



However, that doesn't mean that WD is necessarily superior. Go look at the reviews for any make/model over 2TB...and the reviews are a total minefield. It seems that quality control is just... totally lackluster at best.

I want to get a 4TB Black or two, but even those.. the reviews don't look all that promising. Looked at Seagate 4TBs, too.. also a minefield.

As far as what you should be using in that NAS machine.. WD Red, because they are meant for NAS setups. And Purple, despite what somebody said previously.. surveillance servers don't do long, sequential writes, they do tons of tiny writes. I've managed a Toshiba Surveillix server before, and the written-data is crazy-fragmented. Each 10gb partition was 98% fragmented and had like 800,000 fragments (according to Defraggler).

As far as which colors do what and so forth.. I did a bunch of research about a year ago and read through a lot of whitepapers and so forth to generate this comparison list:


Hope that helps. Personally, I would only ever do black or yellow. There is also orange, and that is 15k RPM, with 10^18 error rate and 1.6M hours for MTBF. But it doesn't have high capacity.

The Western Digital drive died because of the high heat it was exposed to. It is really dead, I tried it in my docking station, it's not recognized as a drive. Not even as an unformatted one.

The company I bought the Shuttle NAS and one of the drives from has only contacted me once last week and since that time it's total silence again. Not a real client friendly company. Once my Windows system is back up&running, I'll go email them again and demand they do something. Warranty on the drives being up or not, they hadn't yet have 300K hours on them (the one surviving has 13.3K hours), they certainly shouldn't have had to live in a 90 degree centigrade box. So in my opinion it's a Shuttle error, they should pay me for two new drives. They can also repair the NAS, put a new fan in and such, but they're not getting away with it.

And if I had put it someplace stupid, like in a cupboard, then yes that may account for extra heat. But it wasn't. It was sitting where the new NAS sits as well, on the floor, besides a box I built for my modem, router and telephone to sit on and all their cables to sit under. It's in front of the door to the balcony. Not in the sun.

I had also set the old NAS to run the fan always, even though that gave extra noise. If I set it to run always and it isn't at all, well... Shuttle says the NAS keeps an eye on that and that it will shut down if the temps go above 62C, it definitely didn't do that. The SMART readout on the drive proves that.

---

Good luck in dealing with them.
I do hope you get some satisfaction.
It certainly seems like you have enough factual ammunition if you would have to threaten with a lawsuit of some kind to bring them around.

Not a day too late I am doing this. The last reboot after AOMEI Partition Assistant Home Editor finished copying over this problem drive to the new drive has made things worse.

Loading Windows is extremely slow. It took just 15 minutes to get from login screen to the desktop. Guess I should be glad it loaded...
Windows Event Viewer can no longer load. Its snap-in tells me something went wrong and it has to unload. OK.
Loading any program costs minutes.

I'll be turning off now, and will just change the drives over tomorrow (take out the 1TB drive, put the 2TB in its place) and reboot from that. And then go sit through a Windows repair installation. Yay!

Well, ain't got anything else to do either, can't go anywhere with the fever and the sneezes.

edit: this copying took 5 hours. Better than the NAS, which is slated to go until 7am. And then I'll have to do that again, because the present array rebuild is done from the damaged WD drive onto one of the Seagates. Eventually I'll have to rebuild the array onto both of the Seagates.

---

It's my first time I run a RAID1, Chris, so I don't really know. I now see that the syncing took till 7 a.m. and that the actual repairing is now taking place with no numbers on it showing how long this will take. I'll dive into the documentation next.

Meanwhilst I changed over to my SSHD and am running the broken Windows off of that. Tried a reboot and indeed, again it's throwing away broken file segments and indexes, indicating a problem with the file system, not with the hard drive.

To be sure I am running Partition Assistant over the old HDD, doing a surface scan first.
Also -again- tried the other hard drive out of the NAS, because this program can read drives that have no normal partitioning record. But nope, it's still dead. Doesn't even start up. Although it did feel heavy on the connector side when I took it out of the docking station again. Anyone know what that is? I've felt that on other drives in there as well, when taking them out that they seem to be heavily magnetic, wanting to be pointing down.

That Gembird docking station is one of the best things I bought in a long time. And while it seems to be only USB 2.0, last night's copying over of data went at a sustained 35 megabytes per second (the max for USB 2.0 as it seems). I did stick it in one of my USB 3.0 outputs and may try to get the 3.0 docking station now, seeing how it has a max of 400MB/sec. :)

Okay, on to repair install Windows. If you haven't heard from me by the end of the day, call in tech support.

Edit1:
It's fighting me on all sides. The upgrade process stopped three times already telling me I needed to reboot and restart the process.
edit2: Make that four times. I could just in time post the above edit, before the system rebooted. Still not installing Windows. And I surely hope it's not secretly installing Windows 10. ;-)
edit3: not looking good. It's been stuck at fetching 172703 out of 10693723 files for just over an hour now. I feel a bad four letter word coming up.

---

There are two kinds of people in the world:
1) Those who have restored a full backup
2) Those who spout bad four letter words

---

A full backup is of no use here, because of the massive amount of programs and games I have installed. Why do you think I do a repair install while maintaining my user settings and not just a reinstall of Windows?

I also know now what is happening: I had to choose the Upgrade with Windows Update option as else it didn't want to start the installation, and that is slowing everything down to a crawl because Microsoft is only giving Windows 10 users a high speed connection for Windows Updates, whereas users of Vista, 7, 8 and 8.1 get these at the slowest possible connections.

The total progress is at 45% now and has been that for the past hour and a half.

I'm lucky I have two monitors attached, so I still can do things via the second monitor, including monitor (ha!) the progress on disk in windows explorer.

edit: this is going to take days. Progress 46%.

---

Got to progress 47% and I just killed the 'upgrade process'. At this rate it's going to take a further 4.5 days before it's at 100%. Windows Updates downloads come in at an extremely slow pace.

I'll go put the Windows partition of the old drive back and work with that. See if I have other options. Better a wrecked Windows than having to surrender my PC to a 5 day 24 hour upgrade process.

---

See if I have other options.

There are "other options"...

See if I have other options.

There are "other options"...

Well, then tell me, oh guru.
But don't mention Linux, please. Not interested. Tried Ubuntu again not too long ago, still don't like it. Not ready to jump through all kinds of hoops to get all my games, old and new, to work on that. Even wine can only do so much.

As for one separate machine for games, Chris, this is it. The only game on the TV server is an ancient version of ESO from the time that Holly's system was taking up space on the dining room table, catching dust for close to 9 months, due to her having given up on trying to rebuild it. (And I wasn't allowed to).

sfc /scannow is still jumping out on 60%, so will go try that in safe mode.

---

See if I have other options.

There are "other options"...

Well, then tell me, oh guru.
But don't mention Linux, please.

Jord, i hope the sfc works (i hope it's not some intermittent hardware fault messing up the file system), but trying to recover a sick OS disk from a sick OS disk is tough especially if the file system from which recovery is happening is not solid. This is true for all OS's.

Seems to me MS are making no secrets of making WinNot10 user's life increasingly difficult, and that will only get worse as time passes.

i'm no guru on these matters, but one option is invest in a SSD if you don't have one for a new system disk, as this will be a good investment, then prove all the hardware is fine and then work on recovering that system disk (as a non-system disk).

See i didn't mention the obvious option...

...but trying to recover a sick OS disk from a sick OS disk is tough

But I wasn't trying to do that. I've only copied over the old drive's contents (5 partitions spanning 1TB) onto the new hybrid drive (2TB) and changed over to using that drive this morning. This was to make sure that the problems I had were not with the drive itself, but with the file system.

I've just ended a full CHKDSK /R /F on the C: partition, that took an hour and a half. Got a clean bill of health.

i'm no guru on these matters, but one option is invest in a SSD if you don't have one for a new system disk, as this will be a good investment

I don't like SSDs, as I find them too small for the money you throw at them. Which is why I bought an SSHD, which is a hard drive with 8GB of NAND memory, in which it will store the most used files. This way operations speed up enormously, and Windows startup can go from multiple minutes to within 10 seconds (or so Seagate says, and with Windows 8).

I have already noticed speed ups, as I have some giant directories that now load instantly.

then prove all the hardware is fine and then work on recovering that system disk (as a non-system disk).

All the hardware is working fine.
This Windows has been in operation since 19 July 2010. That's quite a bit of time for Windows, to have run that long without a reinstall.

So, tomorrow I'll continue.

---

A full backup is of no use here, because of the massive amount of programs and games I have installed.

The number of tears shed rises as the square of the number of bytes not backed up.

Once you get things back to whatever normal you can, get an el-chepo drive and backup that system partition. Unplug the drive and put it in a closet. Say every two or three months, plug it back in and make that backup.

With the drive powered down and sitting still it won't degrade nearly as fast, so an el-chepo will be fine.

Oh, and if M$ ever forces 10, wouldn't it be really nice to have a working 7 to go back to?

---

Once you get things back to whatever normal you can, get an el-chepo drive and backup that system partition.

But that's the thing, that's not very useful in the case of having many games and programs installed. Because a lot of those games and programs are installed on the same hard drive as the one the system partition is on.

So this would only help if something were to happen to the system partition alone, not when something happens to any of the other partitions, not when something happens to the whole hard drive.

My Windows was installed on 19 July 2010. It's been running fine for all up to - checking when I first mentioned its problems - half December of last year. In the mean time this system is one with 3 individual HDDs, with 1,695.4 gigabyte of data between them.

The el-cheapo drive is likely to fail first, because it's got the most data to store and the most write sessions to undergo. And then at 2TB to lose all the backup data...

Still... a failing file system isn't something that's happening from one day in December 2015 to the next one. That's something that slowly builds up. And thus something that one would've been backing up then.

For do you know all the tells and tales of a failing file system and what you should look out for and when that first started? How do you know for sure yours isn't starting to go ever so slightly corrupt at this time? And since you're giving the el-cheapo drive backing up advice, you're one who follows that own advice - what has that done to your backups lately?

---

Once you get things back to whatever normal you can, get an el-chepo drive and backup that system partition.

But that's the thing, that's not very useful in the case of having many games and programs installed. Because a lot of those games and programs are installed on the same hard drive as the one the system partition is on.

Well, if you don't save applications where they belong, there is no hope for you. After all you aren't using the protection against allowing a executable file on a data partition. Oh wait, you aren't on a *nix.

The el-cheapo drive is likely to fail first, because it's got the most data to store and the most write sessions to undergo. And then at 2TB to lose all the backup data...

As above, if you don't have applications, which don't change often so they don't write to the backup often ...
You also missed the put it in a closet. If you were a business this would be the copy in the fire safe. A business would also have an offsite copy.

Still... a failing file system isn't something that's happening from one day in December 2015 to the next one. That's something that slowly builds up. And thus something that one would've been backing up then.

For do you know all the tells and tales of a failing file system and what you should look out for and when that first started? How do you know for sure yours isn't starting to go ever so slightly corrupt at this time? And since you're giving the el-cheapo drive backing up advice, you're one who follows that own advice - what has that done to your backups lately?

You have heard of incremental backups haven't you? The type where you can pick the date of the file you want to restore? Do you know how the backup program decides a file is changed? Consider how a filesystem corrupts.

---

Well, if you don't save applications where they belong, there is no hope for you.

Where do you think I save my applications?
Where do they belong?

As above, if you don't have applications, which don't change often so they don't write to the backup often

The Elder Scrolls Online has a new version of itself out every month, month and a half. That's usually multiple large downloads amounting to a total of 5 gigabyte of downloads. When you now install the game clean, you'll first have to download the latest 3 updates, amounting to 15 GB of data, that then has to be installed separately over each other.

And that's only one game that updates quite often.
GTA V has a new update out every month.
There hasn't been a big game out yet the past 5 years that hasn't got an immediate full length update within 24 hours of release.

You also missed the put it in a closet.

No, I did not miss that. You said "Say every two or three months, plug it back in and make that backup.
You missed I was talking about the full amount of data I have on all my drives, that changes quite a bit over time, with stuff being installed, other stuff being removed, new stuff being downloaded etc. I was talking about the backing up of everything, that 1.6TB amount of data.

I've just lost almost 7TB of data on the failed RAID, with lots of it not in backup because the NAS was the backup. You cannot expect of users to have closets full of backup hard drives with backups of everything on their systems when normally a NAS shouldn't have a problem keeping all that data.

Even with the 300K hours MTBF of the WD Green drives that we had in there, if the NAS' fan hadn't stopped working causing the drives to heat up to over 90 degrees centigrade, and the NAS failing to shut down because it didn't read that amount of heat, the WD Green drives would've easily run for 12,500 days.

The failed WD Green had just over 13.5K hours on time on it, that's 4.5% of the total time it should've been able to survive.

If you were a business this would be the copy in the fire safe. A business would also have an offsite copy.

But I am not a business, am I now?
A business also has most of the work done saved on the server, so you only have to backup one machine, and at that incremental because not all of the work people saved has to be backed up every day, only what they added that day.

Having a user computer versus a big business server is totally incomparable.

You have heard of incremental backups haven't you? The type where you can pick the date of the file you want to restore? Do you know how the backup program decides a file is changed? Consider how a filesystem corrupts.

I have heard of incremental backups, but other than that they only backup newer files/copies of themselves, I don't know much about them.

Doesn't matter what I say about what my intention was, doesn't matter what I all explained in this thread, I should've 5.5 years ago started making backups of my Windows file system only, and made an incremental one every 2 to 3 months and I should have had a backup program that was intelligent enough that it would have warned me as soon as the first time came after a Windows Update that the computer rebooted and started throwing away indexes. Have I missed anything?

Probably that I don't know what date exactly this started happening.
Probably that it has happened way way way way way way way before at a reboot e.g. after a Windows Update and the next checking drive for consistency, and that I wasn't at the keyboard to see that.
Probably...

Ah never mind, the thing I had the problem with -the actual in-place upgrade installation of Windows- is not mentioned anywhere in your unhelpful comments, so why am I even trying?

---

..

...but trying to recover a sick OS disk from a sick OS disk is tough

But I wasn't trying to do that. I've only copied over the old drive's contents (5 partitions spanning 1TB) onto the new hybrid drive (2TB) and changed over to using that drive this morning. This was to make sure that the problems I had were not with the drive itself, but with the file system.

I wasn't clear, if the OS executables, libraries or kernel is damaged, using it to recover (itself) is tough (even if the file system is good)

I've just ended a full CHKDSK /R /F on the C: partition, that took an hour and a half. Got a clean bill of health.

What was the sfc result?

i'm no guru on these matters, but one option is invest in a SSD if you don't have one for a new system disk, as this will be a good investment

I don't like SSDs, as I find them too small for the money you throw at them. Which is why I bought an SSHD, which is a hard drive with 8GB of NAND memory, in which it will store the most used files. This way operations speed up enormously, and Windows startup can go from multiple minutes to within 10 seconds (or so Seagate says, and with Windows 8).

I don't need to restart that often, but 2 seconds is my usual boot time with an SSD with the unmentionable option. I run the VMs if i need them off a 1TB HDD drive they take between 30 and 50 seconds for Windows.

I have already noticed speed ups, as I have some giant directories that now load instantly.

then prove all the hardware is fine and then work on recovering that system disk (as a non-system disk).

All the hardware is working fine.
This Windows has been in operation since 19 July 2010. That's quite a bit of time for Windows, to have run that long without a reinstall.

good luck ... and yes, a repurchase every five years is recommended by the supplier ;) You probably have a half a TB of recovery files now to help you recover too.

Just spent 5 hours getting my system back, after I bought a 3 PC/2 year license for Bitdefender Antivirus Plus. Where it installed without problems on my other two systems, on my private one it gives only trouble. For as soon as I reboot to get the service to run, upon login I have two black screens (2 monitors). Explorer.exe keeps on crashing, no matter what I do.

Trouble is, without explorer.exe you cannot uninstall the offending program, because everything, from Add Remove Programs to checking the directory with Windows Explorer uses the explorer.exe shell.

So I resorted in putting back a restore point, something that this program doesn't make either before it installs itself. It's a bit icky already anyway, as it will uninstall Spywareblaster when it finds it.
But if you only have it mentioned in the registry but not actually installed due to having restored your system to an earlier restore point, at which time the SWB was still installed... the Bitdefender installer will silently exit after giving you a message it cannot find SWB.

Which is irritating, as prior to this one message there is a very slow download happening. I've seen that one 4 times now. Slow? 25 minutes.

So... it would seem I now really have to fix this Windows before I can install AVP16. Tomorrow then.

---

So... it would seem I now really have to fix this Windows before I can install AVP16. Tomorrow then.

Good luck Jord, if all else fails - "Nuke it from orbit, it is the only safe option"