Respect !!!

[Indiana]

Hey, I still have one of those old Quantum LPS270 drives running (I think it's like 1992 or so).
I guess the decrease in hardware prices overall also led to a decrease in manufacturing quality.

[Technoid]

Has anyone thought of the fact that the GXP's where the first really mainstream 7200rpm disks??
Computer stores sold a s**t load of them to people that had only had 5400 and 4200 drives!!!

And I bet that a lot of the "self destructed" ibm GXP's is because of overheating!
(or as in my covorkers case: plugging the airhole!!)

[CHHAS]

Hehe, I still have a Quantum LPS210 running in my parents' rig, those things just won't die.

[Cobos]

Actually all I've heard about the IBM HDD problems IS overheating. Or rather the newer IBM disks are more sensitive to heat than other manufacturer's disks. Thus when they got stuck in hot cases the disks get a bad rep, and of course as technoid said since they've sold a big lot of them you get lots of compliants

Cobos

[Indiana]

The Quantum LPS series really has to be one of the most reliable things ever made.

And I've had one IBM GXP drive die even though I do have a HD cooler!
That tells one thing: either the thing is getting too hot in normal operation or they f*cked up with their quality control. You know a HD is [/I]supposed to[/I] live for several years even without a seperate fan...

I NEVER cared about cooling my LPS back then and it was way hot in my Amiga with the PPC604e@200Mhz, the (uncooled, not even passively!) 68060@50MHz, a Permedia2 based gfx-card and two Harddrives in a small desktop.

[Voyager3D]

Well didn't have any problems with my IBM drives..
I have one 30Gb 75GXP in my gaming rig in a SilentDrive, so i bet it is getting pretty hot. But i have modded a slotcooler to cool it a little (at 7 volt). I have a good and silent cooling in my AOpen tower, it's 33 degrees C right now with 26 degrees C room temp.

In my server i have one 30Gb 75GXP and one 40Gb GV or somthing (IBM's 5400rpm disk). The server has been running for almost 2 years now, without any problems... PSU, CPU and 80mm air intake fan are all running at 7 volt. The disks are placed right infront of air intake.

[fav]

I'll trust my servers HD noone but IBM.

Just keep away from IBM DTLA version if you over-anxious. I heard from some guys they have problems with 'em; so be careful.

The rest series from IBM are superb.

I myself had no problems with IBM DTLA 30GB and 40GB drives nor with older models.

[The workstation machine I have in use currently have IBM DTLA-304050, IBM IC35L040AVER07-7 {check those AVER series!} inside; FUJI drives even not got recognized by genuine Intel chipset/BIOS for P4 845DW mainboard), don't know why].

I will keep my eyes on Western Digital HDDs because they have almost same specs as IBM drives; I suspect that they are use same technology and R&D in production.

Yes, IBM drives get sometimes hotter than competitors but it have inner mechanism to compensate temp overheat (others have such an option too).

Good flow of air will not hurt but is not mandatory.

And I do REALLY hate when last week a latest super-puper Seagate Cheetah SCSI drive in our accounting server got S.M.A.R.T warning and died in a week in constantly controlled humidity environment at a 18.5C conditioned server room *brrr*.

[Wombat]

Yes, IBM drives get sometimes hotter than competitors but it have inner mechanism to compensate temp overheat

What mechanism? The click of death?

[WayneHu]

Originally posted by Maggi


my personal choice would have been the 80GB from Western Digital WD800JB ... 8MB cache, 8.9ms, 7200rpm

dunno, why my sysad chose the IBMs over my choice, but I guess it was some sort of special offer or maybe even a bundled price.

I am waiting for your demonstration of Ultra-Monster Parhelia!

Last year I had more than 5 IBM HDDs. But 3 of them were RMAed due to my fatal action..

I also have two WD800JB for the 24-hr online computer and they work fine for most cases and do not disturb my sleep. But I found most IDE disk sub-systems do not perform well with more than 200000 files in a single NTFS volume. This is not a joke. I have more than 260000 files in a single volume.

Finally I found 3ware Escalade 6410 + RAID 1 system is probably the cheapest cost to deal with so many files with acceptable performance. I attached three WD1000BB on this controller and do the copy action between the single drive and the RAID 1 volume. The CPU utilization and the console's response is unbelivable.

If I put more than 1 HDD just on single ICH2, FastTrak 66, or HPT 372 controller, the console's response speed is somewhat slowed down with more than 2 HDDs' concurrent access. Without 3ware controller, I have to put each IDE HDD on different onboard/Addon controllers to avoid their cheap design problem. An old Ultra2 SCSI controller can even do better jobs with more than 1 HDD if compared with those funny firmware IDE RAID controllers...

If the random access and large small files access are compared, WD800JB still cannot beat the WD Enterprise 10K HD which I put it on my office machine. The problem of 10K rpm HDD is the noise. As a result, I can just use it in office with a lot of background noise to compensate it, not the quiet home bedroom.

[fav]

Wombat: What mechanism? The click of death?

Aha, for better understanding you need to put a link to a page on the IBM site where different cases of broken mechanics can be heard in person.

Click of death is often refered to iomega drives and wrongly promoted further to IBM drives without backin up by any trusted statistics (link, anyone?)

(Atleast Seagate will not let me hear how and why our Cheetah HDD resulted a horrible server death and a a half-hour for reinstallation and re-backuping.)

http://www.cd-info.com/CDIC/Technology/CD-R/Premastering/Factors.html#therm

Disclaimer: my goal is not put a flame into this discussion nor incline anyone to buy IBM disks. My goal is to share my 5 year experience operating a server farm and assembling computers for fun and profit. I'll gladly provide any information I have. Yes, I'll stick to IBM/Hitachi drives as I do in relation with Matrox productions. For me the IBM drives shortcomings (FUD, price) are overweighted by the advantages (stability in my environment, support, performance and compatibility).

[Wombat]

IBM drives have their own click of death. I know, I finally heard my own 75GXP do it after 1-1/2 years of faithfull errorless service. I was copying its contents to a new array, when it gave out after about 25GB of transfer. You want trusted statistics, hit storagereview. Something closer to home? We've had threads here about people's failing GXP drives.

[fav]

I am off the discussion. Only a rhetorical question: do you have a rack of faulted HDDs (died nearby), say 100 pieces of 'em?

I do, collected different kinds of. IBM ones are not something exceptional.

Scale matters when talking about statistics.

My respects.

[Enak]

Just my experience

Just my experience...

I have seen very few HDDs fail which is probably a good thing.

But here are the stats:

IBM 1x failed new 3x failed in use (under 1 year old)
Maxtor 1x failed in use (2 years old)
Seagate 1x smart error - discontinued use (3+ years old)
Western Digital 1x fail in use (5 years old)

I have also noticed that the few IBM branded machines our IT dept have purchased at work contain Maxtor HDDs!

I am currently running 4x seagate barracuda IVs in RAID 0 and 2 others on seperate machines... These drives are the quietest I have ever used and seem pretty robust and good performers.

I am currently awaiting an RMA from IBM for the latest death.

[az]

I'm running a 60 GXP and just recently CPUCool gave me a message that some smart value decreased from 100 to 99... I think it was data errors or something. I did overclock slightly at the time, don't do anymore.

Can anyone tell me if this is a sign of the HD dying, or if it is due to the FSB running at 112, therefore (or therefor?) overclocking PCI, AGP, etc.?

AZ

[fav]

http://forums.murc.ws/showthread.php?s=&threadid=34838