Slashdot Mirror

← Back to Stories (view on slashdot.org)

SCSI vs. IDE In The Real World

Posted by timothy on from the safe-to-go-back-in-the-water dept.

An anonymous reader writes "Gerard Beekmans has a really good comparison of the speeds of IDE and SCSI drives up over on devchannel.org. Should help put an end to the myth of IDE erasing SCSI's speed advantage." Note that Beekmans' test handicaps the SCSI disk a bit, with interesting results. (DevChannel, like Slashdot, is part of OSDN.)

14 of 586 comments (clear)

  1. IDE w/ 2meg cache? by Anonymous Coward · · Score: 1, Interesting

    Why didn't he test one with 8 megs? Or ones of rougly the same size? Or ones with roughly the same anything? He also could've tested a newer Serial ATA drive. Heck, for the price of SCSI, you can build a nice RAID with multiple IDE drives and win back lots of speed. This review is a very big mess.

    1. Re:IDE w/ 2meg cache? by ericman31 · · Score: 2, Interesting

      Pieroxy wrote: Dude: 7 minutes vs 28 seconds. That's more than 1:14!!!

      Dude: This comparison leaves so much out that it is completely meaningless.

      • No mention is made of data layout, I would say that the "benchmark" author didn't bother to ensure that he had the exact same data layout and filesystem layout between the two disks.
      • Were the files clean, or fragmented? Was one disk clean and the other fragmented?
      • Were these the only disks in each system? Or was some of the other I/O mentioned hitting against a different spindle and/or controller?
      • More spindles, smaller disks on the spindles, will give better performance than one spindle (or a few) with a larger disk. I'm betting that he added the SCSI drive to his server and much of the I/O activity on the system was hitting a different disk, possibly a different controller as well.
      • What drivers is he using and what operating system? Is it exactly the same, down to the kernel version on both systems? Did he use a generic IDE driver on one and the vendor's highly optimized and tuned SCSI driver on the other?
      • Are both systems basic PC in nature? Or is one a laptop and the other a low end server?

      I have actually been part of benchmarks that had such a wide disparity in performance. But we could back up our results with rigorous benchmark standards and results. The benchmarks involved several different database engines on the exact same server hardware, storage subsystem and operating system. All system parameters, data load processes, operating system optimizations, DBMS tuning and queries were carefully documented and reviewed by independent SME's. Benchmark results were documented and reviewed after being repeated multiple times for each system. At the end of it all we had a benchmark that was acceptable to our customers, the engineers and the PHB's.

      This so called benchmark would be laughed out the door. SCSI generally performs better for I/O intensive scenarios and that can be proven with appropriate benchmarking. But this "benchmark" was not rigorous or thorough and proved nothing.

      --
      In my universe I'm perfectly normal, it's not my fault you don't live in my universe.
  2. scsi and laptops by KhanAFur · · Score: 3, Interesting

    This article show that scsi drive have a considerable advantage over the same spindle speed of ide drives. Laptops tend to have slower drives. Has anyone considered using scsi drives in laptops?

    Does anyone know fo laptops that use scsi drives?

    -Mary

  3. You get what you pay for. by Sheetrock · · Score: 4, Interesting
    Besides the speed advantage, SCSI drives also typically last two to three times longer than their IDE counterparts, and generally go through more rigorous testing.

    Tape drives are like this, too. They look the same, they act about the same during the write process, but the cheapie drives that come with some servers will fail to reread the tapes if they're reused as constantly as they are in most businesses (who, on average, reuse the same weekly tapes for a full year or more!). Better to put the money into a DLTtape solution than to rely on what's bundled with the server.

    --

    Try not. Do or do not, there is no try.
    -- Dr. Spock, stardate 2822-3.




    1. Re:You get what you pay for. by BWJones · · Score: 2, Interesting

      Besides the speed advantage, SCSI drives also typically last two to three times longer than their IDE counterparts, and generally go through more rigorous testing.

      This is absolutely true. Apple used to ship all SCSI drives in their machines and I still have twelve year old Macs that have never had a hard drive failure. The new IDE drives however are a different story. That said, Apple appears to do more quality testing on their hard drives in that since Apple started shipping Macs with IDE drives, I have had two failures. Compare this with Wintel boxes like Dells where I have had close to ten IDE drive failures.

      The other issue that folks should know is that if you are doing any work that is truly disc intensive (like photoshop or scientific computing) then SCSI has historically been the protocol of choice with much faster speeds possible with SCSI. For instance, I have insisted on 10-15k RPM drives for my work and they are much faster than even the fastest IDE devices. This is starting to change with fast SATA drives however and I am looking forward to some new options with the G5.

      --
      Visit Jonesblog and say hello.
    2. Re:You get what you pay for. by Naeleros · · Score: 2, Interesting

      And why wouldn't Western Digital just make contracts with their customers for certain quantities of drives meeting certain specifications at certain prices, all of which could vary from customer to customer, be it Dell, Apple, or whomever?

      The reason for this has to do with "Just In Time Manufacturing" (no inventory) & low margins. Its not permissible to have inventory because of JITM. Its not feasible to have multiple assembly lines for the same process (excepting if you need the production capacity because of order volume) because of low margins. And, its not reasonable to keep 'resetting' your main production line for a different customer's procedures (imagine you're in the middle of a several day process of changing over when you get a call from your biggest customer saying he needs 16 hrs of parts TOMORROW... it happens all the time...keep those lines ready).

      Its just far simpler to find a common QA level that everyone can agree to. If you have a customer that demands a certain high level of quality on a product line.. then that's what you produce (if his volume is high enough to warrant) for all. You can certainly have varied product lines (which probably span multiple assembly lines). Those varied product lines almost certainly do have varying degrees of QA/testing.

      What I'm asserting is that you don't make 2 products that are exactly alike in every respect except the level of quality that you use for your customers (for big markets like HDDs... if you work out of your garage.. its no biggie). If you have an 'assembly line' that is 'better'.. then you make a new 'product line' and charge more than you charge for the other one. (I'm also saying that Dell/Apple use the same line of HDD's...now that Apple has dropped SCSI as standard.)

  4. Not an accurate test by pagercam2 · · Score: 2, Interesting

    The document lists one and only one case. I don't doubt that SCSI has performance benfits that is pretty well known. I've always wondered why they don't upgrade IDE with a better command set much like SCSI, well they haven't they just increase the clock speed and offer better buffering. So there is a valid case for a comparison between SCSI and IDE. This review does one and only one test which proves that SCSI wins on one test this is not a good article. He reads one and only one file. The real question is how well IDE and SCSI operate under real multi-treaded OS conditions. Slashdot editors should be rejecting this article in favour of one with a real indepth analysis. SCSI will win but not for the reasons listed in this article.

  5. Re:SCSI vs. IDE: Same experiences by Zocalo · · Score: 2, Interesting
    Here's some anecdotal evidence for you. Many moons ago, when CD ROMs were still a rare beast, I had two models of the same dual speed drive on my desk, one IDE and one SCSI. These were state of the art drives at the time. I had a sample disc with a large Video for Windows file - must have been 10MB at least! On the IDE model playback was jerky and the CPU was maxed at 100%, while on the SCSI version the CPU never got above 50% and playback was smooth as silk.

    The reason for this being that SCSI handles far more of the overhead of managing the disk on the controller than IDE, which left much of the work to the CPU. Of course, this technological gap has narrowed considerably with the evolution of IDE into EIDE and now ATA drives.

    I have to confess, I'm a die hard SCSI fan when I can justify it (although I might be swayed by second generation SATA). While the real world performance gap of SCSI-vs-IDE is long gone, SCSI drives are still synonomous with servers, which usually translates into a more robust product. How much is *your* data worth compared to the SCSI price premium?

    --
    UNIX? They're not even circumcised! Savages!
  6. Holy war? by BenjyD · · Score: 4, Interesting

    What is this? Holy war week on Slashdot? In the last week or so we've had stories on BSD vs Linux, Linux vs Solaris, PHP vs Java, Exchange vs Sendmail , x86 vs PPC and now IDE vs SCSI. All that's missing is Vi vs Emacs and I think we'll have pretty much every major computing disagreement covered.

  7. Cache size by jonpry_oneword · · Score: 1, Interesting

    The hard disk cache size has very little effect in linux boxen since the kernel uses as much memory as possible to cache hard disk data. If it's not found in 100's of MB of ram then it probably isn't in the disks cache, and yes system is ram IS faster than hard disk cache. I would be more interested to know what reorginazational effects copying the 50,000 file directory to the other drive had.

  8. Re:Meaningless.. by tchuladdiass · · Score: 2, Interesting

    I'd like to see, just for kicks, a IDE drive hooked up to a scsi controller via an Acard ide-scsi bridge. Is the speed of scsi mostly from the controller, or from the drive mechanics?

  9. Re:ATTN: Slashdot community by tmark · · Score: 2, Interesting

    If this site really used user-moderation to determine what should show up on the front page, much of the time there wouldn't be much up there. Web sites like these depend on having eyeballs, so the editors need to keep throwing stuff up there. If there's some other more interesting article, people will read those instead. But look at how much time you, and the others who (rightly) hated the article spent reading and then posting about it, and look at how long we spent responding to your posts. And think of all the people who read some of the posts and then just went on without responding. Those constitute a lot of eyeballs. Ever wonder why we keep seeing essentially the same stories about e.g. how Windows/Linux/MacOs is better/worse ? Ever wonder why there are so many dupe stories ? Ever wonder how anyone ever gave Jon Katz a soapbox ? Ever wonder how sometimes the most ridiculous stories/claims get a ton of posts ? Eyeballs are eyeballs, whether they're happy or not. It's not such a bad strategy to purposely put up crappy articles, even if only to make the "angry" readership feel welcome.

    Forget your delusions about this being a site for a "community". It's a business.

  10. SCSI over NFS better than Local ATA for me! by MarcQuadra · · Score: 4, Interesting

    I get better real-world performance from my Ultra-2 SCSI drive _over NFS_ than I do for my local ATA/100, and the SCSI disk itself is about 5 years old.

    That says something.

    --
    "Sometimes, I think Trent just needs a cup of hot chocolate and a blankie." -Tori Amos on Nine Inch Nails
  11. reason it's biased by ShadowRage · · Score: 2, Interesting

    is because he just shelled out $700 on his new drive.
    so, he has to make it out as good to prove to himself what he just bought will beat ide.

    also, not to mention he never stated brand names.
    certain manufacturers are very different.