Become a fan of Slashdot on Facebook

 


Forgot your password?
Close
typodupeerror
×
Data Storage

Disk Drives Explained 132

Posted by CowboyNeal from the stuff-to-read dept.
CowboyRobot writes "Magnetic disk drives are one of those things I usually take for granted without thinking about, but I recently realized how little I understood about how they really work. ACM Queue has an article from their 'Storage' issue titled, 'You Don't Know Jack About Disks', which does a very good job of explaining exactly how magnetic disks have evolved since the 70s and how they work today."
This discussion has been archived. No new comments can be posted.

Disk Drives Explained More

Disk Drives Explained

Comments Filter:
  • It's not too hard for ordinary people to understand (I can understand it and I'm not an expert in hardware), though the article didn't really talk about USB thumb drives, etc.

    What kind of controller interface do they (thumb drives) use? ATA, SCSI or something else?

  • Nintendo Gamecube (Score:3, Interesting)

    by rf0 ( 159958 ) <rghf@fsck.me.uk> on Sunday July 13, 2003 @08:37AM (#6427564) Homepage
    It always intresting to see how things work. Nice little thing to add to this is the way Nintendo do copy protection on their disks (although not scritly on topic). Instead of relying on heavy software encryption they went for a nice simple solution. They spin the CD-Rom the wrong way. As such you need special burners if you want to copy it.

    Now thats a neat idea

    Rus

    • Re:Nintendo Gamecube (Score:5, Informative)

      by hbackert ( 45117 ) on Sunday July 13, 2003 @08:57AM (#6427603) Homepage
      Wrong. Found in many places on the Internet: like here in this GC FQA [google.com]: it's not true that GC disks spin backward. You can check it out youself. Instead the CD starts at the outer part of the disk and the laser then moves towards the inside. Normal CD-ROMs and DVDs start at the inside and go outside (and slow down while doing that).

      • Re:Nintendo Gamecube (Score:1, Interesting)

        by Anonymous Coward
        Actually normal CDs and DVDs speed up when moving from inside to outside (as the angular velocity increases along with the distance from the center). Now, what I want to know is ... if you are doing a bit-by-bit copy, what difference would it make if the discs were meant to be read forward, backwards, or sideways?

        • Re:Nintendo Gamecube (Score:2, Informative)

          by Anonymous Coward
          There is no reason why the angular velocity should increase when the laser is pointed further away from the center of roation. The rate at which information is read is R * V where V is the angular velocity, and R is the distance from the center where the laser points to. The CD/DVD drive must slow down (less V) its spin as it starts to read furhter away from the center (more R) because the hardware can process information (R * V) only so fast.
      • I love people that start their posts out with "Wrong!". I was instantly convinced!
      • From what I understand, the Sega(/MS) Dreamcast had a similar method. The data is written from the outside inward. You can pick up a random Dreamcast disc and see the obvious difference.
        • Nope. The Sega Dreamcast discs (Gigabyte Discs) have two sections - a high density and a low density section. The low density section is normal CD and can have audio tracks, data tracks, whatever stored on it and can be accessed with a normal CD-ROM drive. The high-density section is normal CD but is written with the data in a tighter spiral making it unreadable to computer CD-ROMS drives. This also has the unfortunate effect of making them very succeptable to death by scratching. These Dreamcast will
      • Ofcourse it's my head that spins backwards, due to all the this overload of detail, heeeellllp :-)

  • Funniest /.'ed message (Score:3, Funny)

    by Bob Wehadababyitsabo ( 629809 ) on Sunday July 13, 2003 @08:50AM (#6427586)
    "Fatal error: Call to undefined function: message_die() in db/db.php on line 88"

    So... anybody got a mirror?

  • A bit more history (Score:5, Interesting)

    by Mostly a lurker ( 634878 ) on Sunday July 13, 2003 @08:51AM (#6427590)
    Although I found the article interesting in terms of the modern developments in PC hard drives, it is a little misleading concerning the overall situation in the 1970s and 1980s.

    To take the IBM mainframe example he quotes: yes, IBM originally used a CKD (count-key-data) architecture and this was still preferred in the late 1970s for highest performance applications. However, in the last 1970s, IBM already provided FBA (fixed block architecture) disk drives such as the 3370. These moved intelligence of disk geometry into the disk controller and were quite easy to program.

    Other mainframe and minicomputer manufacturers had innovative schemes during the early 1980s.

    • Re:A bit more history (Score:4, Insightful)

      by Rick.C ( 626083 ) on Sunday July 13, 2003 @09:23AM (#6427642)
      When you consider that the average mainframes of the early '70s had around 512K (yes, "K") of memory (the big ones had a meg or two), you can understand the need to conserve memory. Most programs ran in 60K regions. You just didn't have room for a lot of large data buffers. CKD format allowed you to write whatever size records made sense for your application.

      The real beauty of CKD was the "K" or "key" field. If you wrote data blocks with keys, you could then ask the disk controller to search for a given key while your program was executing other code. The controller would find the matching record, read it into storage and interrupt when it was done!

      Nowadays most mainframe DASD is really RAID-1 or RAID-5 SCSI arrays that emulate CKD under the covers. With gobs of RAM and the introduction of "dataspaces", the usefulness of CKD is debatable, but like other legacy interfaces, CKD will be a long time dying.
      • When you consider that the average mainframes of the early '70s had around 512K (yes, "K") of memory (the big ones had a meg or two), you can understand the need to conserve memory.

        Indeed, and actually 512K was a pretty large mainframe in the early '70s. I remember working with a 370/115 with only 64K and a 370/145 with originally 256K. The 370/145 was running VM/370 with DOS/VS, MVS and VS1 guests, as well as CMS users. As you can imagine, performance was not stellar!

    • Re:A bit more history (Score:2, Informative)

      by grigori ( 676336 )
      Another thing about CKD is that early disks were unreliable and you needed to make sure you read the location you thought you were reading (like for instance if seek calibration was off), so by putting a countfield containing cyl/head/record in front of each data field you got self-identifying data. If the read CCW said gimme a particular record and the count field you actually read didnt match you could recover by reissuing the read or otherwise hiccuping
    • The other thing he failed to mention (or I failed to notice) is that mainframes had ordered seek queuing since at least the 80s. Furthermore, the operating system could balance the advantage of ordering the queue to minimize seeks with other factors such as task priority, which is something the disk drive can't do.

      I'm not saying that it is a bad idea to do this in hardware, but it is just one more thing that was invented long ago on mainframes and is now "new" for PCs.

      I'm still convinced that in the long

  • Magentic disk drives are bad for environment (Score:5, Insightful)

    by Krapangor ( 533950 ) on Sunday July 13, 2003 @09:17AM (#6427633) Homepage
    The magenic layers contain very large amounts of chrome-oxides and other heavy metals and rare elements. This makes disk drives a huge problem in the disposal of old computers.
    Furthermore the rare element production takes often place in very anarchic countries like Kongo or Liberia. Usually warlords and local terrorists use the money from the disk drives rare elements to finance their blood raids and terrorship.
    That's btw the reason why the US were setting up Kabila in Kongo. This guy was killed, but only because the French were more clever.

    So, instead of this old technology which is going to be phased out in 5 years anyway, you should use more modern flash/ram disks and DVDs for data storage, just for moralic reasons.
    Think about it: If you refuse to buy bananas or big name brands because of the cruel, inhumane exploitation of the third-world workers, then you should do the same in IT and avoid disk drives.

    • Unfortunately, we're still quite a ways from developing solid-state drives cheaply that can finally compete against today's hard drives.

      I think within the next 15 years we will finally see the breakthrough that will essentially turn our primary storage into physically-removeable cartridges of solid-state non-volatile memory storing over 250 GB of data on a cartridge somewhat smaller than the physical dimensions of a 3.5" floppy drive. And unlike today's non-volatile memory, the new solid-state memory can b
      • The delay in booting an OS is not caused from the data transfer of HDs. Look at linux booting and you will understand. There are lots of devices that need configuration and that takes time, for reasons unknown to me tho.

        Really, why does configuring non-mechanical devices so long? Why does the bios check take so awfuly long? Shouldn't a mobo that has a bus running at 133MHz actualy be up in milliseconds?
        • You'd think that. Except many devices have timeouts that go upto seconds. So when trying to config [e.g.. guessing]...

          Also it wouldn't be impressive if there weren't alot of churning to effectively turn on a computer.

          What you really have to question is why you have your mobo do the PNP setup in a matter of a second then linux still takes about 20 seconds to boot [thereabouts I've never really timed it].

          Tom
        • My theory is poorly written BIOSes. We have computers at work (Dells, notably) which go from power on to booting the disk in about one second. Sure, it skips things like memory tests, but BIOS memory tests are usually pathetic anyway. I've had bad memory that passes the BIOS check with flying colors but throws up tons of errors when you test with memtest86.

          Instant POST is exceptionally handy on laptops, where every second POSTing can be wasted battery life in many cases.

      • Part of the problem is people want 10GB/sec memory.

        Which causes heat and requires alot of expensive memory. Heck I would settle for 50GB of 5MB/sec memory which is all my IDE harddrive can write anyways.

        Of course 5MB/sec wouldn't be popular [even though you could effectively make it with ram from the early 90s]. Which is why you lie, it isn't 5MB/sec its 41943040 bits/sec. Now that's impressive!

        Tom
        • Tom,

          I don't think the solid-state drive I suggested is going to run at the same speed as main system RAM! =)

          What I am thinking of is something more like the memory used on CompactFlash and SD memory cards, only with vastly improved durability and 250 GB capacity. Because you don't have to go through the mechanical process of physically writing and reading data on a spinning disk, such a drive could load the OS is a small fraction of the time it needs now even with today's fastest ATA-133 or Serial ATA har
          • The problem with such memories is that they are not durable.

            A smarter idea would be to have a high capacity li-ion battery backup that recharges when the computer is on. Eventually the battery will die which means making it hot-swappable [e.g. while the machine is on] would prove useful.

            My computer is on 99% of the time so the battery essentially would never be used [except for power outages and the like].

            In the end a battery backed solution would prove more practical for durability [e.g. you can re-wri
      • While we're still using hard-drives, though, things may yet change beyond recognition.

        The other day I came across this discussion [acmqueue.org] between Jim Gray (MS Research) and David Patterson (Pardee Prof. at UC Berkeley) suggesting among other things that if the disks get much bigger (Terabytes) they're likely to be serial-access rather than random-access (the return of mag-tape, but flatter, because that way you could read a 20TB disk in a day instead of a year), and the return of sneakernet (for terabytes of data,

    • Re:Magentic disk drives are bad for environment (Score:1, Funny)

      by Anonymous Coward
      Wow, is this true, or the usual net conspiracy thoery?

      The only reason the parent was modded "Insightful" was that he spelled Congo with a 'K'.
      Remember, alternate spellings always mean that the writer knows the 'real' story.
    • A LOT of mining, of every sort, takes place in underdeveloped countries. Check out, frex, the issues involved with bauxite mining in the 3rd world (bauxite is the ore from which aluminum is extracted, using vast amounts of electricity in the process. Don't DVDs use an aluminum layer? Doesn't that DVD drive have an aluminum frame?)

      Take away the mining industry, and what do you have? A lot of now-unemployed miners with NO other work available, and reduced cash flow in a country that's already strapped, thus
    • This being Slashdot I felt compelled to play the roll of grammar policeman (despite the fact that I know my own grammar to be lacking).

      Unless english is not your native tongue, someone claiming to have superior intelligence (as your tagline clearly makes this claim) should take pains to write clearly.

      This would include not making up new words where others work perfectly well, I.E. "terrorship" instead of "terrorism" and "moralic" instead of "moral". Further, phrases such as "more clever" can be reduced
    • Mensa member, beware of the high IQ

      beware of the high Ego :-)

      To be serious I mean that, Mensa are a load of elitist loosers, if you want friends & people who like you for who you are not for some irrelevant parameter of your being, then loose the Mensa loosers.

  • Relevant Link... (Score:5, Informative)

    by henele ( 574362 ) on Sunday July 13, 2003 @09:22AM (#6427641) Homepage
    There are probably lots of articles on the subject that can be bought up with a quick search, but my favourite broad explanation site has their definition here [howstuffworks.com].
    • If they really new how stuff works they wouldn't be sticking that mouse in their head, owch [howstuffworks.com] don't they know how bad a mouse can be for their brain, :-), repeat after me: I will not be putting the mouse inside my cranium.

  • All you want to know about (Score:5, Informative)

    by Cee ( 22717 ) on Sunday July 13, 2003 @09:24AM (#6427643)
    Well, there's an interesting article about hard drives here [storagereview.com]. 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 [storagereview.com] anymore? =) Or, that in fact, there are holes in the hard drive [storagereview.com], so it can "breathe".

  • Referenced PDF (Score:5, Informative)

    by Helmholtz ( 2715 ) on Sunday July 13, 2003 @09:27AM (#6427654) Homepage

    On page 6 (ATA versus SCSI) Mr. Anderson (insert matrix joke here) references "ATA versus SCSI: More Than an Interface," by Dave Anderson, Erik Riedel, and Jim Dykes.

    The pdf can he had here: ATA_vs_SCSI [gutenpress.org]

    I thought it was quite an interesting read, and an excellent companion piece to the "You Don't Know Jack about Disks" article.

  • Well... (Score:5, Funny)

    by mraymer ( 516227 ) <mraymer@nOsPaM.centurytel.net> on Sunday July 13, 2003 @10:32AM (#6427871) Homepage Journal
    I think acmqueue.org just learned something new about hard disks. Namely, that they turn into a pile of warm liquid goo after a severe slashdotting...

  • Aha! (Score:3, Funny)

    by bythescruff ( 522831 ) <tim.beattie@ntHO ... minus herbivore> on Sunday July 13, 2003 @10:44AM (#6427911) Homepage

    Fatal error: Call to undefined function: message_die() in db/db.php on line 88

    Yup, my hard drive did that a few weeks ago...

  • That server's gotten /'d, so here's a mirror [earlham.edu].

  • Text of article (Score:3, Informative)

    by onomatomania ( 598947 ) on Sunday July 13, 2003 @12:20PM (#6428323)
    Here are direct links to the figure images:

    Figure 1 [acmqueue.org]
    Figure 2 [acmqueue.org]
    Figure 3 [acmqueue.org]
    Figure 4 [acmqueue.org]
    Figure 5 [acmqueue.org]
    Figure 6 [acmqueue.org]
    Figure 7 [acmqueue.org]
    Figure 8 [acmqueue.org]
    Figure 9 [acmqueue.org]

    Magnetic disk drives have been at the heart of computer systems since the early 1960s. They brought not only a significant advantage in processing performance, but also a new level of complexity for programmers. The three-dimensional geometry of a disk drive replaced the simple, linear, address spacetape-based programming model.

    Whatever happened to cylinders and tracks?

    Traditionally, the programmer's working model of disk storage has consisted of a set of uniform cylinders, each with a set of uniform tracks, which in turn hold a fixed number of 512-byte sectors, each with a unique address. The cylinder is made up of concentric circles (or tracks) on each disk platter in a multiplatter drive. Each track is divided up like pie slices into sectors. Because any location in this three-dimensional storage space could be uniquely identified by the cylinder number, head (surface) number, and sector number, this formed the basis for the original programming model for disk drives: cylinder-head-sector access.

    This raises the question: If that is how data is stored on a drive, why don't we still use that as the programming model? The answer is not an easy one but has its roots in the fact that this geometric model endured until the advent of the intelligent inter-faces, SCSI and ATA. [The IBM mainframe world used a slightly different model, allowing tracks to be written with records (blocks) of user-defined length. An individual track could have sectors of different sizes. As one who programmed count key data (CKD) storage, I can attest that it offers the application wonderful flexibility, but the drive design challenges have relegated it to history. Also, a purist might point out that standards etiquette calls for SCSI to use blocks and ATA to use sectors, but I will use these terms interchangeably.]

    Disk-interface protocols implement the programming model for disk drives. The earlier drive interfaces did little more than expose signals to let the host directly manipulate the drive mechanism and initiate a transfer of data at a target location. This put the task of dealing with all the low-level idiosyncrasies peculiar to drives on the programmer charged with developing the firmware or software support.

    The introduction of ATA and SCSI fundamentally changed this. Table 1 describes the migration of intelligence from host to drive in the evolution of the more important interfaces. With these intelligent interface protocols, the task of programming the use of disk drives became much easier. Disk-drive designers also gained a freedom of action needed to design higher-capacity and higher-performance drives. I will look at just how drive designers used this freedom of action in their designs, but it is important first to understand the fundamental goal behind drive design: increasing areal density.

    DAVE ANDERSON, director of strategic planning for Seagate Technology, has more than 20 years of experience in the computer field. His responsibilities include overall strategy for all disk interfaces. He has been involved in the architecture and planning of Fibre Channel since it was first proposed as a disk interface. He was also one of the principal architects of the disk XOR commands that are now a part of the standard
  • Fatal error: Call to undefined function: message_die() in db/db.php on line 88

    Of course... typical php site.
  • Reminds me of 6 of the 12 computer science hours I took last semester; an informative article nonetheless and a good read for those who.... thing about that stuff a lot. I was sure that my programs ran around inside my harddrive fighting bigger three-letter acronymed programs.... oh well, my dreams are stamped out yet again by Dr. Science and his cronies.

  • They get bigger, they get faster, they get quieter. Everyone wants a bigger, faster, quieter one. What else is there to know?

Slashdot Top Deals

Understanding is always the understanding of a smaller problem in relation to a bigger problem. -- P.D. Ouspensky

Close