A Drive Going Bad on me.

[TBoneit]

There is nothing really critical on any drive. I keep things pretty well backed up.

Unfortunately I have 5 drives in this computer and I'm n ot sure what one is going bad. Two Sata, and the rest PATA. I ran Disk Checkup from Passmark. For some reason it is only working on two drives?

I didn't manage to pin down the bad drive (Clack, clack, clack). Probably they are the same two that the Bios reports the SMART on. My guess is C: & D: The other three are on a raid controller although not configured as raid. I'd like to pin down the bad drive and pull it since when it acts up it freezes windows.

Second drive this week, My luck at work. The other one is in a external case.

Any suggestions on testing software that can pull SMART information from all the drives and diagnose them.

Thanks

[Skith]

You can rule out the two that you checked with Passmark, if they passed their tests. Where is the RAID controller located (motherboard or PCI card), does it have it's own BIOS or do all five drives show up in the main BIOS?

If Disk Checkup could only check two, I assume those two drives passed all tests. I would also guess that the three that would not run, would be the three drives running off of the RAID controller.

Also, I am not aware of any tools that will run SMART analsys on external drives connected via USB/firewire.

your best bet is to remove all but the boot and one other drive at a time until you find the problem maker. It might help to remove the drives and lay them on the desk next to the case while they run. Don't touch them or move them while running though.

[lordsmurf]

In my experience, RAID kills non-SCSI hard drives. It might take a few years, but it happens eventually.

Back your stuff up is all I can say.

[Jester700]

In my experience, RAID kills non-SCSI hard drives. It might take a few years, but it happens eventually.

Back your stuff up is all I can say.

Any idea why this would be? I mean, the drive hardware can be the same, so what is there about the controller logic in a SCSI rig that can't be done in PATA or SATA?

[ahhaa]

just guessing, but: IDE HDs were originally designed for intermittent use- load a program, store some data later...
Video usage keeps them constantly running while capturing, rendering, playbacking, etc. With a RAID system, if its set up for max reliability, data is constantly moving from drive to drive- same effect. SCSI drives were designed for servers, and expected to be in continuous use. I heard somewhere that drive heads travel the distance to the moon, tho that doesn't really seem possible!

[TBoneit]

You can rule out the two that you checked with Passmark, if they passed their tests. Where is the RAID controller located (motherboard or PCI card), does it have it's own BIOS or do all five drives show up in the main BIOS?

If Disk Checkup could only check two, I assume those two drives passed all tests. I would also guess that the three that would not run, would be the three drives running off of the RAID controller.

Also, I am not aware of any tools that will run SMART analsys on external drives connected via USB/firewire.

your best bet is to remove all but the boot and one other drive at a time until you find the problem maker. It might help to remove the drives and lay them on the desk next to the case while they run. Don't touch them or move them while running though.

The only two I'm sure of are the Seagates that show up in the Passmark test. The raid controller is built in to the motherboard an ASUS P4P800 Deluxe. The Raid controller uses its own bios, it is not set up as raid, they just run as regular IDE drives. I have started pulling the IDE drives one at a time. Problem is it runs for a while before it locks up. The only time I heard continuous clacking was last night and at that time I couldn't get to drives until I powered down and moved things around.

Wish me luck as a computer that randomly locks up is a headache.

I Have now unhooked all but three drives in the computer C: D: & J:, D: being the XP drive. And relaced the others with External drives that I haven't been using and that have been sitting empty. If I have no problems then I'll pull the others and take them to work for further testing or just replace them with one 300+ Gig drive so I can stop using the externals again since they're slower than internals.

[TBoneit]

Well I've pinned it down to being one of three drives. I actually have more than enough storage for my needs it is just that having had to turn back on external drives to replace these three for testing I can see the difference in speed between a USB2 and a internal drive. Now I need to pull them and bring them into work for rigorous testing to determine what drive is going bad. Especially since one of them is my primary scratch drive and I can see a difference using the usb drive in its place.

[Skith]

In my experience, RAID kills non-SCSI hard drives. It might take a few years, but it happens eventually.

Back your stuff up is all I can say.

Any idea why this would be? I mean, the drive hardware can be the same, so what is there about the controller logic in a SCSI rig that can't be done in PATA or SATA?

Almost completely different. SCSI is designed to a much higher standard. In additon, SCSI capacities are much lower, due the the required reliability. The more densely packed the bits, the more chance of error. SCSI are desinged to take more punishment from 24/7 operation and access read/writes. I have an older 18GB 7200RPM Atlas V SCSI drive (I got it and the controller dirt cheap back around 2000ish). It has been running in one system or another almost 24/7 since then, and never missed a beat. I use it as a thrashing drive. Downloads, tempfiles, music playback/cd rips, and the occasional torrent. It is the scratch disk for photoshop, and several other programs.

SCSI isn't cheap, but for reliability it can't be touched.

[oldandinthe way]

Skith

IMHO opinion your SCSI drive is not fundamentally more reliable than IDE drives. Your SCSI drive is older than IDE drives.

Because your drive is older it has a lower recording density and lower capacity. Any failure will be less likely to be fatal.

As drives have become larger and faster they have to some extent become statistically more unreliable.

However modern IDE drives have more than acceptable specs.

When stepper motor drives were introduced (the full height to half height transition), many conservative systems integrators (myself included) stayed with voice coil drives in the interest of reliability.

When RLL drives were introduced (greater capacity for the asmae number of heads and platters), many integrators stayed with MFM, in the interests of reliability.

Similarly 5.25" vs 3.5" drives. Of course 3.5" drives became the mainstream form factor dominate all but specialty markets.

Many arguments have been made for the SCSI interface - system independance, performance, larger configuration, but I am unaware of any for reliability, which have a valid technical basis.

The mechanical and recording aspects of drives of a comparable generation have usually been identical independant of interface.

If in fact your SCSI drive is more reliable, it is a representation of a general decrease in drive quality, which is not represented in operating specs or mean time between failures (MTBF).

[TBoneit]

I'll buy the fact that the new drives are not as robust as older drives!

When the prices have dropped as much as they have and the capacity has gone up as much as it has, something has to give. I believe what got left behind is sturdiness,

[Skith]

Skith

IMHO opinion your SCSI drive is not fundamentally more reliable than IDE drives. Your SCSI drive is older than IDE drives.

Because your drive is older it has a lower recording density and lower capacity. Any failure will be less likely to be fatal.

As drives have become larger and faster they have to some extent become statistically more unreliable.

Yes, I mentioned above that SCSI have lower capacity, and are thus more reliable (less likely to fail). I suppose you and I view the definition of "reliable" a bit different. I would be willing to bet modern SCSI drives would outlast modern IDE drives if placed in a high stress enterprise enviroment. How can you say they are built the same? Modern SCSI drives reach rotational speeds of +15K RPM. The fastest spinning IDE drive is the WD Raptor at 10K RPM. Clearly there are desing differences in order to compensate for the additional heat, vibratation, and other factors.

-anyways, I won't say anymore on the subjuct as it isn't the main discussion of the OP. I appologise to the OP for this intrusion.

[Wile_E]

Here are some free utilities to monitor SMART. They may or may not work, depending on if they support your motherboard chipset.

HD Tune
http://www.hdtune.com/

SpeedFan
http://www.almico.com/speedfan.php


[/url]

[oldandinthe way]

From the WD website

Device Level Reliability
Hard drive reliability is expressed in number of hours as Mean Time Between Failure (MTBF). It is important to note that Western Digital’s first SATA drive will have the same MTBF (1.2 million hours) as a SCSI drive. Keeping this in mind, also consider that the statistical mean represented by MTBF provides information about a population, but not particular drives. The failure of a particular drive cannot be predicted—hard drive failure is an ever-present reality. Therefore, it is reasonable to expect the reliability of SATA hard drives to approximate that of SCSI drives.

http://wdc.custhelp.com/cgi-bin/wdc.cfg/php/enduser/std_adp.php?p_faqid=942&p_created=...BlPWFuc3dlcnMu

[TBoneit]

Thanks Wile_E.

I recall paying around $1000 for a 1 gig WD drive when they first came out back in the day. I would expect that drive to be better built than the drives of today. Some of the drives I see these days you can feel a steady vibration when they are spinning. they don't seem to be worrying about balancing the platters, shortening bearing life IMHO.

BTW that $1000 drive first one didn't hold up, the replacement outlasted two computers until it got to be to small it was still working. The first one died right away, probably damaged in shiiping is my guess.