Forgot your password?
typodupeerror
Businesses Databases Oracle The Almighty Buck

Oracle Fined For Benchmark Claims 81

Posted by timothy
from the ratchet-down-to-damn-lies dept.
pickens writes "Information Week reports that the Transaction Processing Council, which sets benchmarks for measuring database performance, has fined Oracle $10,000 for Oracle's ads published August 27 and September 3 on the front page of the Wall Street Journal which violate the 'fair use' rules that govern TPC members by 'comparing an existing TPC result to something that does not exist.' The ads said to expect a product announcement on October 14 that would demonstrate that some sort of hybrid Oracle-Sun setup would offer two-digit performance on the TPC-C online transaction processing test compared to IBM's 6 million transaction per minute result on its Power 595 running AIX and DB2. The TPC Council serves as a neutral forum where benchmark results are aired and compared. 'At the time of publication, they didn't have anything' submitted to the council says Michael Majdalany, administrator of the council adding that that Oracle is free to use TPC numbers once it submits an audited result for the Sun-Oracle system. Fines by the TPC are infrequent, with the last action — a $5,000 fine — levied against Microsoft in 2005 for unsupported claims about SQL Server. 'It takes a fairly serious violation to warrant a member being fined,' says Majdalany."
This discussion has been archived. No new comments can be posted.

Oracle Fined For Benchmark Claims

Comments Filter:
  • by neonprimetime (528653) on Wednesday September 30, 2009 @02:35PM (#29597073) Homepage
    ... that this $10,000 fine will cripple Oracle's ability to compete in the future
    • by TheRaven64 (641858) on Wednesday September 30, 2009 @02:41PM (#29597149) Journal
      The five posters below you seem to have all said the same thing, but I'll reply to you:

      The $10K isn't important. The $10K is there so that when a customer asks an IBM (or Microsoft or whoever) representative about Oracle's ad claiming that they can beat IBM's numbers, the IBM sales rep can say 'they were fined for publishing misleading and unsupported numbers. They don't actually have a machine that gives those numbers' and move on.

      • Re: (Score:3, Insightful)

        by Kjella (173770)

        Ding! This is like a $1 fine to normal people, the important part for all the other vendors is they can now say this ad was bogus.

      • Re: (Score:3, Interesting)

        by afidel (530433)
        The problem wasn't really with the numbers, vendors will give you unpublished TPC results frequently but they have to note that they are not certified results. The big problem was publishing the claims very publicly by putting them in a national paper. To me as a consumer the biggest issue is non-certified results almost never have the accompanying audit report which gives the exact config and cost of the solution. Saying that you have a more scalable platform doesn't mean much to most shops if the cost is
        • by FooAtWFU (699187)
          Out of curiosity: Exactly what authority does this three-letter acronym have to fine people for using "uncertified" results, anyway? What makes Oracle obligated to pay? Has Oracle signed contracts or membership agreements with them to make them subject to this sort of review? Can they fine me for saying VaporDB is a billion times faster than Oracle?
          • by afidel (530433) on Thursday October 01, 2009 @12:50AM (#29601997)
            Yes, TPC is a trade organization with all of the major players being members. They actually offer a very valuable service which is independently audited and certified results on vendor neutral benchmarks which fairly accurately reflect real workloads. Any changes from the base configuration of the OS or product have to be documented as does the hardware configuration used and the price of the solution. (including discounts, but you better believe I'm asking for at least their published discount)

            I find that it's a good place to start when looking for solutions as it gives you a good idea of product performance, pricing level for your needed performance, etc. It also can give you some ideas on what tweaks can bring big performance gains as the vendors generally have their best people helping with these type of published benchmarks so the documented config changes can be very useful if they aren't pure benchmark fluff.
      • That'll show 'em!!

      • The $10K is there so that when a customer asks an IBM (or Microsoft or whoever) representative about Oracle's ad claiming that they can beat IBM's numbers, the IBM sales rep can say 'they were fined for publishing misleading and unsupported numbers.

        But if IBM say that Oracle were fined 100K, would they get fined a million?

        Hey, I think I've got an idea to solve the financial crisis!

    • ...that the $10,000 fine was a lot cheaper than the front page ad in the WSJ.
    • by Jeian (409916)

      I dunno why this is modded troll.

      Sarcastic and biting, but still a valid point...

    • I am almost certain ... that this $10,000 fine will cripple Oracle's ability to compete in the future



      I would be inclined to diagnose atrophy of moderator's irony gland.
    • by JPeMu (942971)
      $10k? Surely it means $10k per CPU-C (Copy Published of Unverifiable Claim) ;)

      I'll get my coat...
  • by Thornburg (264444) on Wednesday September 30, 2009 @02:36PM (#29597089)

    Even if Oracle knew they would be fined $10,000 it was probably still well worth the cost of the fine + the cost of the ad. Not to mention that receiving the fine has gotten them the front page of Slashdot and probably lots of other tech sites as well.

    Value for money, 10 Grand was a steal.

    • Re: (Score:3, Insightful)

      by Anonymous Coward

      Jesus, since when is commenting that some company got value out of the they paid for a fine make someone a troll?

      It's not like I said "oooh, ouch, $10k will really put the hurt on Oracle" like the other 12 people who posted in the first two minutes.

      (Yeah, that's my comment that I'm replying to as AC... just in case people are still blowing mod points by marking everything in sight as Troll).

      • Re: (Score:2, Insightful)

        by Nyall (646782)

        I stopped taking the moderation here seriously a long time ago. Its not worth the brain cells.

        • Re: (Score:1, Insightful)

          by Anonymous Coward

          I'm tempted to mod you insightful.

        • by MrMista_B (891430)

          I get 5 new mod points every 24 hours whether I use them or not. Which is awesome, but I don't even come here all that often.

    • They are exactly right. The only thing that matters is that the ad might register with some purchasing manager. Accuracy be damned.

      Meanwhile the honest db developer/supporter is irreparably harmed. They can't possibly defend it.

    • Re: (Score:3, Informative)

      by illumin8 (148082)

      Value for money, 10 Grand was a steal.

      Funny thing: I wanted to get a quote for the Sun/Oracle Database Machine that they are advertising as having these ungodly performance numbers. You know how Oracle licenses their database software per CPU? Well, they have extended their ungodly license to their Exadata storage with a $10,000 per HARD DRIVE license. Yes, that's correct. Oracle takes standard Intel based Sun servers, loads them up with SATA drives, and charges you a $10,000 per spindle license fee

      • by bertok (226922)

        Value for money, 10 Grand was a steal.

        Funny thing: I wanted to get a quote for the Sun/Oracle Database Machine that they are advertising as having these ungodly performance numbers. You know how Oracle licenses their database software per CPU? Well, they have extended their ungodly license to their Exadata storage with a $10,000 per HARD DRIVE license. Yes, that's correct. Oracle takes standard Intel based Sun servers, loads them up with SATA drives, and charges you a $10,000 per spindle license fee to store data on them. This is their business model.

        Does anyone know of any open source alternatives to Exadata? The architecture looks appealing from a performance standpoint: Standard Intel servers with SATA drives connected to a 40 gigabit Infiniband fabric and serving data to Oracle servers, but I'm not willing to pay $10K per spindle to license my storage in the same way that Oracle licenses their database software.

        Look at Sun Thumpers: they're 48-disk storage servers that use ZFS to RAID data. Use iSCSI and high-end NICs to connect to Oracle. You can get multi-port 10GbE NICs for a reasonable cost these days, and a lot of vendors include iSCSI offload.

        Some 10GbE switches now have very low latency, comparable to Infiniband. Or, if you've already got Infiniband infrastructure, just keep using that.

        For performance, pack the Thumpers with RAM (I think 128GB+ is doable), and use the ARC cache feature of SUN Solaris in com

    • by alexo (9335)

      According to Wikipedia, Oracle's operating income is US$8.32e+09.
      If we compare it to a nice salary of, say, $100K, the $10K fine would be equivalent to fining the hypothetical person 12 cents.

      Where to I sign up?

  • $10k is a marketing expense... not a penalty. It won't change anything.

    • Re: (Score:3, Insightful)

      by PitaBred (632671)
      Yeah, but that all their competitors will be able to market by saying "Oracle got fined for lying about benchmark claims"? That's priceless.
  • huh? (Score:5, Funny)

    by Nyall (646782) on Wednesday September 30, 2009 @02:44PM (#29597195) Homepage

    I'm an embedded engineer so could someone tell me: is two digit performance better or worse than 6 million per minute ?

  • That much? (Score:1, Redundant)

    by sjbe (173966)

    has fined Oracle $10,000...

    Wow. How will Oracle ever come up with that kind of cash? Larry Elison might have to sell a seat cushion from one of his yachts to cover the bill.

  • by Seakip18 (1106315) on Wednesday September 30, 2009 @02:45PM (#29597231) Journal

    The $10k fine isn't what Oracle is really being hit with. Depending on how serious the TPC is taken by customers or after MS or IBM run their market-o-tron speak on the actual news, this is an easy to use market strategy against Oracle.

    A-queue-the-show....

    PHB: Why should I go with IBM over other solutions like Oracle?
    Marketing guy: Wait, you're serious? Oracle? The company that can't even get benchmarks right, let alone compare them?
    PHB: What are you talking about?
    Marketing guy: Heh. Oracle's benchmarks are being decried by the industry*. You can't trust those snakes that pose as reputable sellers of database products.
    PHB: Oh teh noes! I hate snakes! Let me buy snake repellent from you now!
    *market-o-tron recommends not giving specifics but make broad generalizations.

    • by eln (21727)
      I thought everyone already knew that benchmarks in marketing materials were pure fiction.
  • The summary, by calling it "some sort of hybrid Oracle Sun thing" implies the product itself doesn't exist, when in fact the issue is that the results of a TPC test on the product were not vetted by TPC (or maybe the test wasn't even conducted yet, it's not clear) before Oracle decided to advertise them. The "some sort of hybrid Oracle Sun thing" is Exadata 2, and it's a real product.
    • by mikeee (137160)

      More likely an Exadata 3, or 2S, or something.

      Exadata 1 was just preinstalled Oracle on HP hardware. Exadata 2 is essentially the same but with Sun x86 gear replacing the HP. Neither of these have the sort of numbers Oracle is promising here, and they've all but said it's Sparc. I'd guess it's a cluster of Sun T5440 (4 socket, 32 core, 256 thread) Sparc servers (which have very respectable throughput, despite marginal single-thread performance) with Oracle RAC and some sort of Sun disk array, probably wi

  • by PingPongBoy (303994) on Wednesday September 30, 2009 @02:56PM (#29597387)

    Basically, Oracle takes a calculated risk of a bad reputation vs. making buyers hold off on purchases from competitors for a short time. My guess is that Oracle will be able to produce something living up to the hype.

    Why not have a little excitement and see if a competitor will match what Oracle is predicting? I can bet that in the labs a lot of products do a lot better in some areas than the released versions. Maybe IBM can loosen the reins and run with Oracle.

    Fun in the capitalist sun. Or is that Sun?

    • It hurts them more than it helps. Certain industries are notorious for bad publicity or even bad rumors being worse than any amount of great benchmarks. We take the benchmarks with a grain of salt, but manipulative misleading ads tell us that they don't have the performance. Both the database and enterprise storage industry take these sorts of performance claims seriously and watch them careful. We also take data loss issues so seriously (and with such paranoia) that if a vendor has any data loss, almost no

  • by Archfeld (6757) * <treboreel@live.com> on Wednesday September 30, 2009 @03:12PM (#29597577) Journal

    The group that is responsible for selecting DB's for the large scale customers Oracle is after is a relatively small select(*)(pun intended) group of people. I attend a national DB conference every year for going on 10 now and I see the same people. Word like this gets out and around. $10k seems like nothing but the fact of them getting fined gets to the people responsible for the product selection and HURTS A LOT more than a $10K fine. I assure you I will be harrassing the Oracle engineers and sales people about this and ensuring my boss, the one who signs the checks is WELL aware of the issue so he can squeeze oracle like the slightly rotten grape it really is....

    • Re: (Score:3, Interesting)

      by Major Blud (789630)
      As a DBA, I agree with this. I'm constantly being hounded by salesmen touting their OLTP numbers. I, for one, am glad that this was brought to my attention; it will give me more solid information to use when countering their arguments, as well as forcing them to watch their advertised stats move closely.
  • I can't believe nobody has brought up this nugget..

    Oracle had hoped to complete the purchase of Sun by the end of August, but a secondary investigation by the by European Commission announced Sept. 3 into restraint of trade has delayed its completion until January or later. Oracle Chairman Larry Ellison has said in an interview that Sun is losing $100 million a month as deal winds down to completion.

    Seems like a lot until you realize they have almost $3B in the bank.

  • This ad was on the back of the Economist 3 or 4 weeks ago... can they be sued for that too? I felt the ad was very very misleading and meant to trick all those MBA guys who want to pretend like they know what they are doing.
  • >> a $5,000 fine -- levied against Microsoft in 2005 for unsupported claims about SQL Server.

    Gee that fine will have made Microsoft change their ways. NOT.

    By being so soft on members the TPC have totally undermined their own credability.

  • Everyone said Oracle had no experience in the hardware and systems business. This pretty much proves they can run with the big boys.

  • So now IBM can say, "Not only did Oracle get fined for an ad in the Wall Street Journal for untrue facts, their fine was double what Microsoft received for a similar infraction."

    Double the punishment a known and convicted monopolizer and business abuser received.. Hmm...

  • Effect on MySQL? (Score:5, Interesting)

    by devleopard (317515) on Wednesday September 30, 2009 @06:21PM (#29599639) Homepage

    A bit misleading, but Microsoft can now say,

    "Looking to implement MySQL? The corporate parent of MySQL was fined for publishing untrue statements about database performance in the Wall Street Journal"

  • Gear the punishment to sales. For example, in Europe the traffic fines are related to the person's income. So the head of Nokia got a speeding ticket for 12 million dollars [beataspeedingfine.com].

    In this case estimate how many sales were affected by this lie, and make the fine equal to the estimated profit on those sales. Then this type of problem would never happen again.

    Wait, never mind, I forgot we are talking about a company based in the United States of Corruption.
    • ....or gear the punishment to the amount of the performance claim. An eight-digit fine (that's two digits plus the other six digits) would make them take notice more than the five-digit fine actually levied.
    • by afabbro (33948)

      Gear the punishment to sales. For example, in Europe the traffic fines are related to the person's income.

      LOL! Not so much...

      First, it's only Finland. Second, there's a minimum floor, which is set rather painful for the poor. So gee, looking a little closer, it's not quite so fair, is it? If you're poor, you still pay a minimum ($106 as I recall), but if you're rich, the sky's the limit. They talked about making it completely proportional, but then the government would lose income...

      Finland's law is simply a reach for money by bureaucrats. I await your comment about The Finnish Republic of Corruption.

  • Every Oracle agreement includes language that Thou Shalt Not Publish Benchmarks. And they're really serious about it - if you use Oracle DB, you can't publish any benchmarks. I wonder if this isn't someone's payback
  • Great, I need a good fight to watch over the next few months. I'm sure Larry Ellison won't disappoint.

"Love is an ideal thing, marriage a real thing; a confusion of the real with the ideal never goes unpunished." -- Goethe

Working...