Backup fail

So, here's my setup:
I have 2 1.5TB internal SATA drives and 3TB WD MyBook external drive I use for backup.
Both 1.5TB drives are just mirrors of each other. I used to have windows mirror set up, but after couple of times the mirror broke, I've decided to simply synchronize both drives manually.
We have recently moved and after unpacking and plugging the computer back in, everything seemed to be in place and working.
Today, I noticed that my main of the 1.5TB drives isn't showing in Explorer. MyBook shows as unformatted space.
After restart and running CHKDSK, my main drive showed up. I checked the secondary and discovered that some of the files are missing, compared to main.
I've started my sync software. In the middle of the operation, my main drive disappeared again.
In Disk Management, it shows as failed.
Reactivate volume doesn't help. Only after a restart I see that drive for a few minutes, then it disappears.
MyBook looks completely wiped out.

Looks like I'll need to replace either both of the drives, although my suspicion is that SATA controller on MB is failing, so maybe I need to replace MB first.

Question 1: How's your backup is organized, since both of my backups seemed to fail. Luckily, I was able to recover the main drive temporarily, but long enough to copy data off it.

Question 2: What kind of drives do you use (external, internal, brand, capacity)?

Two drives failing at the same time is incredibly unlikely. Try them on another PC.

My main drive is mirrored, and I take incremental backups occasionally. Everything is offsite in two locations, and important stuff that's not too large goes online to mozy and crashplan.

I haven't lost data - yet. Remember all hard drives fail, the only question is when.

I suspect either MB or SATA controller.
Those drives were acting funny for the last couple of years, but never this serious.

The minute a drive acts funny replace it, or track down the problem!

ya i would suspect a problem with another hardware component rather then both the hard drives. highly unlikely both of them would fail at the same time.

test them in another computer to see if they are recognized properly there.

in regards to your question.

my current setup is a 1tb and a 500gb.

do manual backups of important data.

currently building a nas system to offload everything to one central location and have it set up with a redundant raid.

Aleness wrote in post #16044355
We have recently moved and after unpacking and plugging the computer back in, everything seemed to be in place and working.

Open the case up, and go over every connection making sure it's properly plugged in.
I know a few people (including myself) who have had SATA cables, power cables etc, come loose during transport.

tim wrote in post #16044436
Two drives failing at the same time is incredibly unlikely. Try them on another PC.

Except when manufactured close in time and having the same uptime and operating in the same environment. Perfectly produced drives would normally have a very long expected life and the fail time could be expected to differ a lot. But lots of times, the disks aren't perfectly manufactured. So there may be a conceptual manufacturing error giving both drives almost identical fail times, or failing for the same stimuli. Lots of raid owners have noticed this - one disk fail, and during the extra load when synchronizing a replacement disk, one or more of the original disks will fail.

A single batch of components with a design defect could have multiple disks fail just because of a single, quite tiny, voltage spike that should have been supported by the hardware.

A number of times, there are also direct firmware errors that makes multiple disks fail for similar situations. Like Seagate who sent out a firmware that "intentionally" bricked disks in case some errors where seen - just to make sure the hardware would be fit enough for recovery. Ooop1 - that bricking could trig for the wrong reasons, bricking a perfectly healthy disk. Ooops2 - that bricking made it extremely expensive to send in the disk for data recovery. While a limping disk would have allowed a user to read out most (or all) data for free.

So no - it can never every be assumed to be unlikely for multiple disks to fail at the same time, or close after each other.

Moppie wrote in post #16047727
Open the case up, and go over every connection making sure it's properly plugged in.
I know a few people (including myself) who have had SATA cables, power cables etc, come loose during transport.

SATA cables are definitely prone to come lose when equipment is moved or bumped into - which is a reason why there are special SATA cables with an extra lock on the connectors.

Aleness wrote in post #16044355
I've started my sync software. In the middle of the operation, my main drive disappeared again.

When something is wrong and you don't feel comfortable about a source drive, you should never (!) perform a sync to one of your normal destination disks. Do the file copying to an empty directory tree somewhere.

If the source drive misbehaves, you could manage to copy a new - but damaged - file, and overwrite a potentially slightly older - but correct - file on the target disk.

Always operate on the assumption that the original disk is totally toast, and you rmirror copy is now your original. Would you allow programs to write to your originals before you have managed yet another copy of these files?

On another issue - never use mirroring as backup. It isn't. Mirroring is a concept intended for additional availability, i.e. to allow you to have a directly accessible directory tree available if your main drive fails.

But mirroring doesn't just copy what is good. It also copies errors and oopses. If a virus destroys a number of files on the master disk, then the mirror will be happy to duplicate this to your "backup" disk(s).

So run synchronization software if your goal is improved availability.

But run backup software if your goal is backup. Backup software may not store the files in a way that they are directly usable without you first doing a restore operation. But backup software makes use of versioning, giving you multiple copies of changed files. So you get time to react if a file is trashed, before the previous versions gets removed from your backup sets.

There is a middle ground - archiving. If files are known to be never modified, then the backup can be a strict mirroring. But archiving software should not duplicate that file again if it is modified - instead the archiving strategy should raise an alarm that one of your files in the archive have, for some reason, changed. This would allow you to start looking for reasons for corruption - or spot accidental file saves.

And of course - make sure your backup strategy includes off-site storage in one way or another. Fires doesn't care about the number of local mirrors or backups you may have.

Aleness wrote in post #16044355
So, here's my setup:
I have 2 1.5TB internal SATA drives and 3TB WD MyBook external drive I use for backup.
Both 1.5TB drives are just mirrors of each other. I used to have windows mirror set up, but after couple of times the mirror broke, I've decided to simply synchronize both drives manually.
We have recently moved and after unpacking and plugging the computer back in, everything seemed to be in place and working.
Today, I noticed that my main of the 1.5TB drives isn't showing in Explorer. MyBook shows as unformatted space.
After restart and running CHKDSK, my main drive showed up. I checked the secondary and discovered that some of the files are missing, compared to main.
I've started my sync software. In the middle of the operation, my main drive disappeared again.
In Disk Management, it shows as failed.
Reactivate volume doesn't help. Only after a restart I see that drive for a few minutes, then it disappears.
MyBook looks completely wiped out.

Looks like I'll need to replace either both of the drives, although my suspicion is that SATA controller on MB is failing, so maybe I need to replace MB first.

Question 1: How's your backup is organized, since both of my backups seemed to fail. Luckily, I was able to recover the main drive temporarily, but long enough to copy data off it.

Question 2: What kind of drives do you use (external, internal, brand, capacity)?

what motherboard do you have? what sata controller do you use?

I have total about 40-50TB of hard drive space at home. Over the years I went though many, many hard drives, probably most experienced person on this forum when it comes to this.

I only buy internal drives, because I want to know what I buy, when you buy external you never know what might be inside.

I have drives from WD, Seagate, Hitachi, Toshiba, Samsung.

WD - ok company, Black series are the best, Blue is second best if your on budget, Red have potential for good raid drive, but had very rough start with drives catching fire. Green = Crap! Never buy green WD, had 5-6 before, 4 failed after few months. Got them RMA'ed and sold off.

Seagate - ok company, in past they had some firmware issues. generally reliable, had more drives from them now currently have 4-5.

Hitachi (Toshiba) - Currently my favorite drives, currently I own 10-12 drives. All their drives work great with Raid. Never hard a single one give me any trouble. Zero compatibility issues with multiply types of controllers. WD recently acquired Hitachi storage division, but had to sell some assets to Toshiba. Right now you can still buy Hitachi drives, and currently Toshiba drives are re-branded Hitachi drives, even the label still says Hitachi.

Samsung - used to excellent drives, but Samsung sold their HDD division to Seagate, and Samsung Hard drives are really hard to find. used to be super reliable drives. Really wish they still would be around.


My hard drives are kept in few location:
In a server
In Sans Digital Hard Drive Enclosure http://www.sansdigital​.com/towerraid-/index.php
In Vantec Hard Drive Dock http://www.vantecusa.c​om/en/product/pro_list​/70

nekrosoft13 wrote in post #16048224
what motherboard do you have? what sata controller do you use?

WD - ok company, Black series are the best, Blue is second best if your on budget, Red have potential for good raid drive, but had very rough start with drives catching fire. Green = Crap! Never buy green WD, had 5-6 before, 4 failed after few months. Got them RMA'ed and sold off.

I have Gigabit MB (can't remember the model number). SATA controller is build-in. Can't comment much on this.

Funny you mentioned your thoughts about different WD drives - the ones that give me trouble are Green! I've started looking at replacement and was considering Red. Can you tell me more about them catching fire?

It isn't so easy to just say what drives - or manufacturers that are good or bad.

IBM did send out tons of excellent drives. Then they had a source code "accident" making people start calling their "DeskStar" line of disks "DeathStar" because of issues with "click of death". Nothing wrong with the hardware, but the firmware made them fail. After the software got fixed, they managed lots of more good disks before being sold to Hitachi.

Seagate also managed many years with excellent disks before they managed a firmware "feature" that made the disks "intentionally" brick. After that fix, they manage to ship good disks again.

WD have released lots of good disks. And also managed to ship firmware optimized for Windows systems, making some of their disks gaining huge amounts of load cycles when used with other operating systems. So a brand new disk could reach the rated number of load cycles within justs months.

And if looking at external disks, almost all manufacturers have managed to release disks that gets overheated - just because we customers don't like hard disk enclosures with fans.

In the end, it really is hard to select a good disk. The disks we buy might be good or crap. History from earlier models from the same manufacturer can't tell us if there are any new goofs released. All we can know is that external enclosures often have temperature issues, and that firmware more often than hardware tends to be the weak spot.

So everyone loves that 3TB drive - that doesn't tell us if the new 4TB drive will also be good or better used as a paper weight. So it often doesn't hurt to try to spread the risks. Maybe one set of backups to Seagate disks and another to WD disks. Or buying disks from separate model lines. Or from significantly different production times.

I had lots of "DeathStar" drives. But zero failures because they where all actively used 24x7 and so not affected by the firmware bug. I did have a number of Seagate drives, and did suffer several failures. The WD firmware issues did bite me before I spotted the crazy SMART-data output and made sure the drives got their configuration tweaked. Only the future will tell which disks I buy now will work well and which will end up well-known as door-stop material.

Red drives are rather new, I haven't bought any yet, when initial batches were released, there were reports that they caught on fire.

Red drives are very similar to green, main difference is that they have build in firmware support for raid.

pwm2 wrote in post #16049236
So it often doesn't hurt to try to spread the risks. Maybe one set of backups to Seagate disks and another to WD disks. Or buying disks from separate model lines. Or from significantly different production times.

thats what I do, I always buy drives from any different manufactures, when I find bad ones, I RMA'ed and sell them off.

In the past 6-8 years, the least amount of trouble I got from Hitachi drives.

Most amount of trouble I got from WD Greens and Seagate when they had firmware issues.

One of the decisive factors was 5 year warranty for WD Black, but looks like Hitachi Ultrastar provides the same 5 year warranty and costs $40 less.
Looks like it might be time for me to make a switch (I've used a lot of WD in the past)