Slashdot Mirror

"You forget the important fact that as the drive DENSITY increases, so does the amount of data read per revolution of the platters"

The _evidence_ of actual transfer rates is more important that your "important fact".

This might be helpful. Select WB99 transfer rate - Begin.

If you have evidence of significantly faster single drives do let me know.

Their IDE RAID card, the SuperTrak SX6000 does REALLY poorly at some tasks. It eats CPU and from mailing lists has a lot of problems recovering from drive failures. For a good comparison to other ATA RAID cards see this storage review writeup on it.

Those are valid points. RAID 1 [or 10 or 01] might make an even better choice for these areas which need fast write speeds.

In case nobody's yet posted the link, Storage Review has a wonderfully detailed section on RAID performance.

Those are valid points. RAID 1 [or 10 or 01] might make an even better choice for these areas which need fast write speeds.

In case nobody's yet posted the link, Storage Review has a wonderfully detailed section on RAID performance.

Why not read a few FAQ entries at StorageReview?

In short, I would probably recommend RAID5 if you have 3+ drives.
RAID5 gives you the most available space while still being redundant. It allows for exactly one hard drive failure.
RAID5's write speed is usually terrible, especially with a small number of drives, but write speed isn't a big deal on my home file server. (Only you know about your needs).

RAID1+0 (NOT RAID 0+1, which is inferior) is great for performance. With 4 drives, you have potentially twice the STR of one drive (writing) and 4 times the STR of one drive reading. Of course, since STR is not important for most IO, this doesn't really effect your end performance much unless you are dealing with linearly reading/writing very large files.
Writing performance will almost certainly be higher than with RAID5.
You do lose quite a lot of space (especially when you use a large number of drives). If you used a 4-drive 1+0 array, you would have the space in two of those individual drives.

RAID1 is nice, and is very reliable, but is impractical with more than two drives unless you are incredibly paranoid. RAID1 simply makes all drives copies of the others, this, you always have as much free space as one drive would have, even if you have ten. If course, you could also handle 9 drive failures and not lose data. RAID1 is fine for 2-drive arrays though.

DO NOT FORGET that RAID is no substitute for regular backups. RAID will not help if your data loss is caused by FS corruption, a cracker, accidentally typing "rm -rf /", etc.

For lowest cost, I would use software RAID, such as Linux's LVM, FreeBSD's Vinum, or whatever Windows has. (RAID5 requires Windows server). (I would not use Windows as the file server myself).
For slightly higher cost, try a Promise controller.
I would avoid Highpoint and Silicon Image controllers. Highpoint, especially, is crap. (but it is very cheap, at least).

If you possibly can, I would recommend a nice 3Ware Escalade controller. Escalades are true hardware RAID cards, unlike Highpoint/SI and most of Promise's cards, and are OS independent and very stable (with certain exceptions for some unlikely configurations).

If you have any questions, you might try the StorageReview forums. There are a number of extremely knowledgeable people there, including engineers and executives-level researchers at hard drive companies. They can give far better advice than I can, I am sure.

By the way, all my comments assume that all drives are the same size. If not, treat all drives as if they are the same size as the smallest drive on the array (unless you are using JBOD, which is not redundant)

Why not read a few FAQ entries at StorageReview?

In short, I would probably recommend RAID5 if you have 3+ drives.
RAID5 gives you the most available space while still being redundant. It allows for exactly one hard drive failure.
RAID5's write speed is usually terrible, especially with a small number of drives, but write speed isn't a big deal on my home file server. (Only you know about your needs).

RAID1+0 (NOT RAID 0+1, which is inferior) is great for performance. With 4 drives, you have potentially twice the STR of one drive (writing) and 4 times the STR of one drive reading. Of course, since STR is not important for most IO, this doesn't really effect your end performance much unless you are dealing with linearly reading/writing very large files.
Writing performance will almost certainly be higher than with RAID5.
You do lose quite a lot of space (especially when you use a large number of drives). If you used a 4-drive 1+0 array, you would have the space in two of those individual drives.

RAID1 is nice, and is very reliable, but is impractical with more than two drives unless you are incredibly paranoid. RAID1 simply makes all drives copies of the others, this, you always have as much free space as one drive would have, even if you have ten. If course, you could also handle 9 drive failures and not lose data. RAID1 is fine for 2-drive arrays though.

DO NOT FORGET that RAID is no substitute for regular backups. RAID will not help if your data loss is caused by FS corruption, a cracker, accidentally typing "rm -rf /", etc.

For lowest cost, I would use software RAID, such as Linux's LVM, FreeBSD's Vinum, or whatever Windows has. (RAID5 requires Windows server). (I would not use Windows as the file server myself).
For slightly higher cost, try a Promise controller.
I would avoid Highpoint and Silicon Image controllers. Highpoint, especially, is crap. (but it is very cheap, at least).

If you possibly can, I would recommend a nice 3Ware Escalade controller. Escalades are true hardware RAID cards, unlike Highpoint/SI and most of Promise's cards, and are OS independent and very stable (with certain exceptions for some unlikely configurations).

If you have any questions, you might try the StorageReview forums. There are a number of extremely knowledgeable people there, including engineers and executives-level researchers at hard drive companies. They can give far better advice than I can, I am sure.

By the way, all my comments assume that all drives are the same size. If not, treat all drives as if they are the same size as the smallest drive on the array (unless you are using JBOD, which is not redundant)

Array Capacity: (Size of Smallest Drive) * (Number of Drives - 1). linky

and

Capacity N-1 Where N is the number of disks, the capacity is N minus 1.
linky.

storage review did a writeup a while ago comparing RAID 0 performace to that of a single drive. more often that not you're better off getting a single, faster drive if you're looking for desktop performance.

21MB/s isnt all that fast. The new WD SATA drives are from 35MB/s to 60MB/s. No, a 5400 drive wont cut it, but any 7200 drive made in the past 2 years should be good. See here and look at "WB99 Disk/Read Transfer Rate - End in MB/Sec".

Note that seek time is not access time, which is the number that really matters.

For example, my (10,000 RPM, Ultra 160 SCSI) Atlas 10K III drive has a measured access time of 7.9 ms, while Fujitsu's MPG3204AH-E (7,200 RPM ATA-100) drive clocks in at 14.8 ms. (These numbers were aquired from storagereview.com, whom I consider the best in the business with disk testing.)

Theoretically, these drives should be 3 ms apart - but there's a 7-ms discrepancy! The Fujitsu drive takes nearly twice as long to access data, when rotational latency only accounts for half that.

Here's a thought:

Keep an idle-quiet hard drive in the box, but don't mount it. Instead, write your raw audio data directly to the drive's device file.

There won't be any seeking, so there won't be any noise. Write raw number of bytes of the total sample to the end of the drive, so you know where your data ends and garbage begins.

Look at a Seagate Barracuda 7200.7 or so; they's practically silent (I have 4 of them; an 80G, 2 160G's and a 200G. My ball bearing 80G Maxtor is probably louder than all of them put together. Other new drives are similarly quiet, of course, but Seagate are still regarded as being the quietest. Check Storage Review :)

Tom's Hardware has nothing to worry about from IBM.

IBM's GXP Deathstar hard drives, as /. regulars are well aware of, are exactly that. Death comes to your data on these drives eventually. Too bad for a large number of customers, it came sooner rather than later.

When the news first broke on these drives, some tech sites came out with the news, and others kept fairly silent. Silence isn't a crime. But continuing to use Deathstars in review gear should be. Why? Because some readers, myself included, used reviews and testing gear examples from Tom's Hardware to build our first computers. Take advice and recommendations from the experts, and you get a better computer, right?

As the current /. story points out, why bite the hand that feeds you advance facts on hardware under ndas, and direct contact with company engineers?

Consumer Reports buys everything they test. With the money that Tom's Hardware has made from advertising on its site (from reader views), they should be doing the same.

Don't take my word for it. Check the dates of when the Deathstar stories first appeared. Then check the hardware reviews on Tom's Hardware. Not just hard drive reviews. Check reviews of other hardware related or dependent upon hard drive speed to get some benchmarks or results. Then see what hard drives are used in the benchmarks, and in the review gear.

While some of their readers went down in flames, others were announcing that the there was a problem, and they continued on as if nothing was wrong. They may have acknowledged the problem in a small story or two iirc (maybe not even that), but they continued using the hard drives in their review gear, without a footnote or warning about them.

Why?

Back in the day of MFM and RLL controllers, the hard drive controller did much of what the drive electronics and firmware do in modern hard drives, that's why you could have MFM or RLL controllers. Hard drives still use RLL encoding today.

Couldn't tell you what "aureal density" means.
That's probably because I can't type. You may want to read this reference for " areal " density, though.

While manufacturers are moving away from heads that touch platters, it is recent and not universal.

Learn about HDDs here

The heads should never touch the platters while the drive is spinning at full speed, (very bad, but it happens, and will cause damage) but the heads rest on the platters when the disk is stopped. The spinning platters drag enough air to lift the heads as they approach full speed. When stopping, the heads are carefully put into the landing zone to park so that there is no danger to the data.

There are exceptions. IBM started parking the heads up and away from the platters on some laptop drives around 1998, so Hitachi inherited the technology. Fujitsu now uses something they call rampload that is probably similar.

This still looks far from universal, unless the drive manufacturers are trying to keep it a secret. These same manufacturers silently redefined a megabyte to boost their capacity numbers by less than 5%. They'd brag about it.

If take apart a disk drive carefully enough, you'll see how the heads rest. Fun way to dispose of a broken HDD too.

Yup, just like Firewire is so well accepted, and SCSI has become a standard feature of all PCs.

Just FYI:

"IEEE-1394 is defined part of the SCSI-3 family of related standards, and was at one point sometimes called "serial SCSI". It is, in fact, a type of SCSI, based on the broad converage of SCSI-3, which goes beyond regular SCSI to cover several similar, "SCSI-like" technologies. In terms of signaling and some aspects of operation, IEEE-1394 really can be thought of as "serial SCSI"."

Until WD improves its QC, I will stay away. In the last year, I've bought Seagate, Maxtor, and Hitachi drives -- all without issue.

The WD Raptor is super sexy, especially in the new 74GB flavor, but I just can't risk another WD drive.

1) Memory hotplugging exists today. It's hardly an insurmountable problem.

2) Even if it were (and didn't already exist) MRAM (Magnetic RAM) is non-volatile.

Still, I have to admit that hard drives have been "scheduled" to be replaced or obsoleted 3 or 4 times now, and every time, they have survived. They are just cheap and versatile and "fast enough", and for applications that want a high sustained transfer rate (STR), they are really quite fast. Fujitsu's latest SCSI drive can handle nearly 80MB/sec sustained for more than half of its capacity. Yo would need a hell of an expensive FLASH controller to outpace that, and FLASH technology is still hampered by a "limited number" of write operations before it dies.

It depends on the exact function of the machine, but it's important to remember that different parts of a disk perform differently.

Why put /boot at the beginning? It only gets accessed once during a run and unless you're rebooting constantly, shouldn't be used very much at all. So, put /boot in the slowest performing part.

How often do you use swap? Depending on the OS and the purpose of the machine, swap might never be used, or when it gets used it's a Bad Thing (e.g. you should have enough RAM to handle 99% of situations). So, swap might not need to be in the best performing part either.

If the machine is a mail server, you might want /var/spool/ or /var/mail or the equivalent in the fastest part. If syslog server, then /var/log gets best performance.

This link shows some stats for older disks. My own tests on more recent disks (seagate 200mb, 8mb cache and wd 250mb,8mb cache ide drives) show similar results - a 2:1 difference from the fastest to slowest sectors of the disk.

Once we start talking about RAID, things get even more complex. Are you sure the same zones are being used from one test run to the next? Your software volume manager might be able to tell you, but hardware raid can be more difficult to tell.

Some posters commented in this thread that fat32 can support up to 2TB in theory, it's just that win2k/xp (perhaps also 98/me?) will not let you create or format such a partition. The problem with a 2TB fat32 partition is that the file allocation table itself will take up 256MB ram (or even 1GB if you want smaller clusters), see storagereview.

Can't FAT32 go up to a terabyte?

Nope:

Link

First if there is a documented section of the command set for changing firmware you use that and only that to change it.

There is no such standard in ATA/ATAPI. See http://www.storagereview.com/guide2000/ref/hdd/if/ ide/.

Second you make the procedure to actually alter the stored firmware complex, so that it is hard to do accidentally, and such that if the procedure is not completed nothing is changed.

That assumes that the device has enough RAM to store the entire image prior to writing it. Many low-cost peripherals do not.

Maybe you don't need to make things quite as complex as for changing the firmware of spacecraft though, since you can at least get physical access to the thing.

When I worked on satellite firmware, it was actually not that dangerous. Unlike consumer electronics, there is hardware that lets you directly write to memory. You don't have to rely on the CPU and firmware in the bird being functional to start with.

When you look at Moore's law for chips and then look at the performance of hard drives you can see that the hard drive is the biggest non inovative mechanism in your computer.

Holographic storage will be the key but until then why don't these hard drive manufacturers try to innovate by adding more than one read/write head on each platter of the hard drive.

Multipe actuators in hard drives

However, since most SCSI hard drives have more cache, higher spin speeds, and lower seek times, they tend to be faster by virtue of *those* attributes. In time I expect SATA RAID will give us all the best of both worlds. Cheaper, more reliable, better performance, and greater capacity per $ doesn't look to be that far off.

Cool, thanks for the link, storageview looks really good, a lot of info there I'd never seen.

Note also that IDE drives in general are "tuned for desktop usage patterns". I'm not entirely sure what that entails, but I suspect it involves a lot of read-ahead caching; single-user systems tend to be actively reading only one or two things at a time. SCSI is tuned for server performance, and the test of "read lots of small files" is probably much closer to a "server" load than to a "desktop" load.

I'd have to disagree. I think logic dictates that "desktop usage" includes opening up your email client, browsing the web, and running small applications. All everyday stuff deals with small files and frequent accesses...which is what you just said SCSI excels at. Granted this is just unknowledgeable reasoning by me, but it seems to make sense...which is why within a year I hope to finally go to SCSI for my OS/applications drive, and keep IDE for storage (video, etc).

Granted, the guy's test was hardly scientific...I'm not basing opinions on it at all. I'm surprised no one has mentioned how easy it is to compare drives at http://www.storagereview.com

Since the market has come to be dominated by the two interface standards, IDE/ATA and SCSI, the question of "which is better" has been bandied about endlessly on the Internet and in other places as well. The simple answer is that neither is better than the other, absolutely. Why? Because if one were clearly superior to the other in every way, the superior one would have taken over the market completely and forced the other one out. The fact that both are in common use (and have been for many years) proves that both have valid reasons for being on the market. Contrast this to the older ST-506/412 interface, for example, which was made obsolete by IDE/ATA since it had no advantages to cause it to persist, and disappeared very quickly.

Whether IDE/ATA or SCSI is better depends on what your needs are, and how much you are willing to spend. In this section I attempt to put the issue in perspective by looking at various performance and quality aspects of the interface and seeing how IDE/ATA and SCSI stack up at each. Of course this will be colored somewhat by my own experience and biases (the entire web site is--that's just life). However, I am attempting here to be as objective as possible in contrasting the two interfaces, and trying to be comprehensive in looking at all of the aspects that are relevant to making a decision.

A primary deciding factor in the SCSI vs. IDE/ATA question is the number of devices you plan to use (or use in the future) in your system. In many respects, IDE/ATA is superior if you are using only one, two or three devices such as hard disks or CD-ROMs. If you are using many devices, say over four, then SCSI is superior to IDE/ATA in several different respects. I should also mention that many times people make the decision about what interface to use based on what interface they began with; few are willing to buy all new devices in making an interface switch. And of course, it's quite possible to use both interfaces in the same system.

Me thinks that someone needs to read up on how MTBF is defined by hard drive vendors. Ohh, and more to the point, how it is defined for SCSI drives as compared to IDE drives (they aren't always the same thing ya know).

For those looking to learn a thing or two today, Storage Review has a nice little explination of MTBF for you all to read. Now, if you do a bit more reading, you can find out just how the definition of "service life" often changes from IDE drives to SCSI drives (for the lazy, most SCSI drives don't include the first 90 days of use in their "service life").

Stop posting stories about single drive failures, they are irrelevant, instead go to Storage Review and check out the reliability database. With the exception of the Deathstars and some problem models from other manufacturers most IDE drives have decent failure rates that generally seem to jibe with the manufacturers MTBF. And if you really care add all your current drives and all future purchases to the reliability survey and go back when they die or go out of service. I have a total of 10 Maxtor's in the reliability database from my personal systems and none of them are dead, still statistically irrelevant but it does help add to the data pool. I also had a batch of ~100 Hitachi laptop HDD's where nearly every one died within a year, now THAT was an obviously bad batch and not statistically insignificant, but most people aren't reporting those kinds of results.

nope. windows xp can only format filesystems as fat32 up to 32gb. however, windows 2000 can format fat32 disks much larger than that (i've never hit the limit). the reason this is intentionally disabled in xp is to discourage the use of fat32 in favor of ntfs on new drives. whether this is due to purposeful compatibility breaking (yes, everything reads fat), or simply a well-intentioned plan to get rid of the ancient fat file system is only known to microsoft.

Also, Anandtech and Tom's have had some reviews in the past.

Personally, I've had bad experiences with my onboard Promise in Win2k... worked fine with 40GB IBM, but when I got a 60GB IBM, vague errors started appearing in event viewer. Promise & Gigabyte did not answer any of my emails. IBM said they normally like Promise, but had no suggestions. I gave up and plugged back in to the Intel controller, and haven't had a single problem since then (over 2 years now).

I've also seen vague errors on Linux when using Promise controllers, and they still don't have an open source driver.

Ive been using this card 24-7 in an dual Athlon system since December and its been great. I chose to do RAID 10 with 4 drives. Rebuilds do kinda suck, but the 3ware desktop manager allows you to adjust the parameters to make general use acceptable. Probably wouldnt want to spool any video while its doing this, but everything else is ok. I believe I paid about $300 then, newegg has for $275. It also does raid 5. Keep in mind they have a newer version out said to be faster. In anyevent, anybody considering raid should read this

Well, there's an interesting article about hard drives here. Watch out, though, it's quite long and it's an easy way to waste some hours... But sure you wanted to know why we don't need to park the hard disk's heads anymore? =) Or, that in fact, there are holes in the hard drive, so it can "breathe".

Well, there's an interesting article about hard drives here. Watch out, though, it's quite long and it's an easy way to waste some hours... But sure you wanted to know why we don't need to park the hard disk's heads anymore? =) Or, that in fact, there are holes in the hard drive, so it can "breathe".

Well, there's an interesting article about hard drives here. Watch out, though, it's quite long and it's an easy way to waste some hours... But sure you wanted to know why we don't need to park the hard disk's heads anymore? =) Or, that in fact, there are holes in the hard drive, so it can "breathe".

Drives are almost never vacuum sealed. See this article from storagereview.com.

Their List of Drives by Heat Output shows the Maxtor DiamondMax 16 (160 GB ATA-133) to be the coolest at 13 degrees Celcius and the next coolest/largest coming in at 200Gb and 19.2 degrees C. I highly reccommend going in and doing a head to head comparison because they break the temp readings down even further. If you're not going solid state, StorageReview can held you make up your mind as to the best drives. (not affiliated w/SR even though i sound like their pimp)

Their List of Drives by Heat Output shows the Maxtor DiamondMax 16 (160 GB ATA-133) to be the coolest at 13 degrees Celcius and the next coolest/largest coming in at 200Gb and 19.2 degrees C. I highly reccommend going in and doing a head to head comparison because they break the temp readings down even further. If you're not going solid state, StorageReview can held you make up your mind as to the best drives. (not affiliated w/SR even though i sound like their pimp)

Okay, If you want to store data on the drive, it's a REALLY bad idea. Incredibly bad, in fact. On current HDDs the fly height of the head is incredibly small. Check out an excellent primer on the subject from Storage Review at http://www.storagereview.com/guide2000/ref/hdd/op/ heads/opHeight.html. The graphic's at the bottom's the important bit. I work with the guys that used to design these things. Unless you have access to a clean room (and even then it's iffy) this is a Bad Idea.

Actually, yes. This would be awesome in corporate situations. StorageReview does this for HDs (registration + submission of data on one HD required to view database). You get stats on the number of people who have them, problems reported, time in use, etc. The bonus for Microsoft is that they can automate the reporting, whereas StorageReview can't (for obvious reasons).

July 2000

April 2003

Seek times have improved by 30%, transfer speeds by over 100%.

July 2000

April 2003

Seek times have improved by 30%, transfer speeds by over 100%.

I agree with your main point. I am a bit suspicious of your alleged 4GB drive from the "early 90's" though. I don't believe 4GB drives were even avaliable until the mid-90's. An interesting chart (that only applies to IBM drives) may be found here.

You can't go by brand alone - at some point every manufacturer has had a line of bad drives.

StorageReview has a Drive Reliability Survey that lists statistics for many drive families. For example, WD 205Bx drives are near the top of the rankings (99th percentile) while the 600Ax is near the bottom (10th percentile).

My current 7500rpm Segate drive makes noticeable amount of noise, this one is even noisier, why can't the drive manufacturers come up with some noise suppression case/jacket for the drives. For my new desktop I would rather go in for 2 low speed (4500rpm) drive in a RAID 0 configuration.
The more recent Seagate and Samsung drives are among the quietest available.
Take a look at the noise state on StorageReview.com's database. You may have an older model, but their new ones are barely audible.

Review of First 10K IDE Drive
Posted by
CowboyNeal
in The Mysterious Future!

from the kicking-the-tires dept.
Sivar writes "StorageReview has a review of the first 10,000 RPM IDE hard drive. Despite the speed that other technologies are improving, this is the first rotational speed increase in almost six years for standard IDE drives." The review is pretty thorough, but also warns to keep in mind that the reviewed unit is only beta hardware.

Review of First 10K IDE Drive
Posted by
CowboyNeal
in The Mysterious Future!

from the kicking-the-tires dept.
Sivar writes "StorageReview has a review of the first 10,000 RPM IDE hard drive. Despite the speed that other technologies are improving, this is the first rotational speed increase in almost six years for standard IDE drives." The review is pretty thorough, but also warns to keep in mind that the reviewed unit is only beta hardware.

One. The 180GXP's outer-zone transfer rates weigh in at a respectable 56.2 MB/sec

Two. The drive more properly act like twins in WB99's transfer rate tests. Here both drives post an outer-zone score of 56.5 MB/sec

Does this mean you will give me two cookies? I like chocolate chip.