DAW Hard Drive Crash Poll - Have your had it happen to you?

[Bob's Mods]

Simple poll. DAWs only please. Have you ever had your DAW harddrive crash? Happen to know what harddrive brand it was? Anybody unlucky enough to have two crashes? About how old was your harddrive when it gave up the ghost? If you operate a two harddrive DAW (and you should) which was it, the application drive or the audio drive?

The idea is to get an sense of how many have had this problem vs those that have not. Hopefully the percentage of users that experience a crash are way less then 2%!

Bob the Mod Guy

 

[dgatwood]

Well, I rarely use my laptop for DAW purposes, but I occasionally do. Its drive died. Brand was Seagate. In the same week, I also had two Seagate desktop drives malfunction fatally (one mechanical, one electronic) in my MythTV setup.

Seagate is now on my do-not-buy list.

I've also had occasional failures of IBM (mostly difficulty spinning up after sitting for a while or bearing noise problems w/o data loss, but I had an 80 GB Deskstar eat itself once). I lost a Maxtor once. I've lost a dozen Quantum drives back when they were still in business.

All hard drives suck. My rule is this: if it is important, keep at least three copies on separate drives. Keep at least one drive off site.



BTW, your question about whether you have a separate system drive or not is largely unimportant. What kills hard drives are three basic things:

1. Temperature outside the safe range.
2. Design flaws.
3. Mechanical impacts (can be cumulative damage to head arms and similar in the case of laptops, too).

Assuming you cool your drive correctly and don't drop your computer, a desktop drive should last forever in the absence of #2. Unfortunately, #2 isn't rare anymore. Within the category of design flaws, there are a few frequent categories:

1. Head failures due to improperly designed parking ramps. This causes damage when the heads are parked (spin down the drive) repeatedly.
2. Bearing lockups/leakage. These issues are exacerbated by heat, and are most commonly seen when a drive first spins up.
3. Faulty air filters. This causes dust to get inside and wreck things.
4. Stupid electronics failures. This includes bad caps, spin chip failures, overwritten track 0 sync bytes due to a buggy drive firmware, etc. These really suck, and there's not much you can do about them.

In short, the best thing you can do to maximize your drive's life is to disable energy conservation and leave it spinning 24x7. Oh, and don't cool the drive too much. The recommended range is 36-47C (97-117 deg. F). According to Google's stats, drives that were running at 27C (87 deg. F) or cooler had twice as high a failure rate as drives within the optimum range.

 

[Bob's Mods]

I have two Seagate Barracuda drives in my system that been running smoothly for many years now. Also, I wonder how if a drive runs too cool it could be a problem? Mine run in a cool case. I can see where heat would be a problem but being to cool mystifies me.

 

[RD423]

It was my audio drive, someone was telling me that I should try changing out the case before I do anything else cause sometimes the externals can lose connection on the inside. Data recovery is expensive and not in my budget. Oh and I think it was a Western Digital external drive just not to sure about it.

 

[Bob's Mods]

I knew of someone who swapped the case on his drive and it worked as the disk was not the problem. Thats a hairy one though because those drives are sealed in a clean room to keep out floating particle matter. This guy got lucky as he had two similar drives. With drives changeing so fast most would have a hard time with this.

 

[dgatwood]

I knew of someone who swapped the case on his drive and it worked as the disk was not the problem. Thats a hairy one though because those drives are sealed in a clean room to keep out floating particle matter. This guy got lucky as he had two similar drives. With drives changeing so fast most would have a hard time with this.

When people talk about swapping the drive case, they aren't (except in the case of a few people who are freaking insane) talking about opening up a hermetically sealed hard drive. They're talking about ripping an internal hard drive out of an external drive case....

Inside an external hard drive, you'll find either a standard desktop or laptop hard drive, a power supply (for desktop-sized drives only), and a bridge board that converts from ATA (or, occasionally, SATA) to some other interface like USB or FireWire.

Moving the drive to a different case is no harder than cracking open a PC, pulling out the drive, and moving it to a different PC. You take out a couple of screws, open up the case, unplug the ATA cable, unplug the power cable, and plug it into the ATA cable and power cable inside the new case (or inside a computer, or...).

 

[suprstar]

I lost a seagate 160gb sata drive once. Secondary data drive, not my OS. I bought it off ebay and it was less than a year old at the time. I keep good backups so I didn't lose any data. No big deal, gave me an opportunity to stick a WAY bigger drive in there, so I'm not complaining!

 

[Farno]

I had a hard drive crash. Man, it sucks when you don't back up data. I don't know what brand it was. It lasted like not even 2 years I think.

 

[Bob's Mods]

When people talk about swapping the drive case, they aren't (except in the case of a few people who are freaking insane) talking about opening up a hermetically sealed hard drive. They're talking about ripping an internal hard drive out of an external drive case....

Inside an external hard drive, you'll find either a standard desktop or laptop hard drive, a power supply (for desktop-sized drives only), and a bridge board that converts from ATA (or, occasionally, SATA) to some other interface like USB or FireWire.

Moving the drive to a different case is no harder than cracking open a PC, pulling out the drive, and moving it to a different PC. You take out a couple of screws, open up the case, unplug the ATA cable, unplug the power cable, and plug it into the ATA cable and power cable inside the new case (or inside a computer, or...).

This guy actually broke open the sealed case and swapped the raw disk. He got very lucky.

 

[fraserhutch]

I've never heard of that being don.e Learn something new everyday.

I've been runnign DAWs since the early 90's I think. I've lost several hard drives. The majority would have been IBM, who are on my do not buy list. I have lost a seagate or two as well.

All drives will fail in time. It has nothing to do with design flaws. Just mechanical entropy. We can increase the MTBFs, but we'll never totally eliminate it. Stuff just wears down and out. Expect it and plan for it.

 

[dgatwood]

This guy actually broke open the sealed case and swapped the raw disk. He got very lucky.

Yes, that's bordering on asylum material. 99% of the time, a board swap will fix the problem unless you have a broken head or something. Did this guy have a broken head or something? Was it one of the ones in the drive?

 

[dgatwood]

All drives will fail in time. It has nothing to do with design flaws. Just mechanical entropy. We can increase the MTBFs, but we'll never totally eliminate it. Stuff just wears down and out. Expect it and plan for it.

Yes, all drives will fail. That said, an awful lot of drives seem to have at least one critical design flaw these days. Frankly, I'm starting to wonder if they are designed to fail with the hope that the majority will fail out of warranty so you'll have to buy a new drive. Could explain why all the drive manufacturers seem to be reducing their warranties lately.... The level of very early (within the first year) drive failures I'm seeing with modern drives is at least an order of magnitude higher than the level of drive failures I saw ten years ago. It's scary how much worse modern drives are, IMHO.

I keep hearing people (particularly in the HD industry) claim that the higher density makes drives more likely to fail. That would be fine if the failures I'm seeing were mostly random bad blocks. They aren't, though. Density changes don't cause heads to break off because the parking ramp is designed wrong. Density changes don't cause click-of-death because the head preamplifiers failed because the lead-free solder joints broke. And so on.

Lead-free solder is to blame for about 90% of the problems these days as best I can tell. I guess it will get better eventually, but... I've about had it with these crappy Winchester designs....

 

[antichef]

I replaced the drive in my macbook (what I use for DAW) with a larger one. While I was at it, I thought I'd be slick and go for a 7200 RPM instead of the 5400 that came with it, thinking I'd get better performance (even though I did not have a noticeable problem with the 5400 one).

Things were OK for a while -- no noticeable performance increase, but I noticed the internal fan in the macbook was running more often and louder. Bad, in my case, since I tend to track in the same room with the computer, being a home recordist and all. I made a few songs on the new drive without backing up, and then one day.... sckiizzikk! After I figured out it was the drive (almost certainly a heat problem), I put the old one back in and booted up fine. I haven't yet fished my projects off the 7200 drive, since OS X won't read the partition, and I don't want to spend $100 on recovery software for one job (anyone know of a free option?)

I got a much bigger 5400 drive and did the swap again, and everything has been copacetic since then.

 

[dgatwood]

Things were OK for a while -- no noticeable performance increase, but I noticed the internal fan in the macbook was running more often and louder.

Part of that may be a lack of RAM causing the HD to run more, but part of it is that 7200 RPM laptop drives are quite a bit hotter. I haven't found it to be too objectionable in mine, but I'm also not doing any significant amount of recording with it....

You might also look into coolbook. Worth a shot to see if it helps with the heat.

 

[Bob's Mods]

Failures seem to be higher than what I would have expected. This is cause for concern. Maybe a RAID system would be best.

 

[dgatwood]

Failures seem to be higher than what I would have expected. This is cause for concern. Maybe a RAID system would be best.

From talking to folks who have tried both, statistically speaking, RAID tends to increase your risk of something going catastrophically wrong (a second drive failing while you are rebuilding the RAID set) unless you don't use any two drives of the same model, manufacturer, production run, or wear level.

Or, as one person put it, if you have one drive of a given model fail at 10,000 hours, the odds are very good that a second drive will fail at 10,000+/-k hours for some small value of k.

Like I said earlier, many of the problems with drives these days are due to design flaws, not abuse, normal wear, or random bad luck. As such, they have a strong tendency to be clustered time-wise. If you have a RAID 5 set, that has data loss written all over it.

Also, don't forget that the MTBF of a RAID set is the MTBF of a single drive divided by the number of drives (or for a mixed set of drives, I -think- it would be the sum of the MTBFs of each drive divided by the square of the number of drives, but don't quote me on that). The point is that the more drives you have, statistically speaking, the less time you will have before something breaks.

Add to that the whole issue of buggy RAID controllers, etc. and you're fairly significantly increasing your risk of losing data by using a RAID, IMHO. No RAID solution can adequately replace a proper backup strategy.... RAID is for performance, not reliability.

 

[rayc]

HDD once, Motherboard once. BOTH were total disasters.
Fortunately I'd learnt to back up by te 2nd. I've still lost some stuff perm.

 

[superspit]

Part of that may be a lack of RAM causing the HD to run more,................

Hi DG...
I find this intriguing......it seems correct.....
I enjoy this concept, never really thought about it before...... I wonder if it's actually true?

Now where's that extra stick of Corsair....oh, and my left sock....

 

[Bob's Mods]

These failure numbers are simply much higher than I expected. I never realized this was such a problem. Its a good heads up for everyone.

 

[mrscruffy]

Seagate HD in a desktop - 2 weeks before I was due to hand all my final work in for my sound engineering course