Oracle Open World: Good News , Bad News

November 3rd, 2015


Looking around Oracle Open World this year, it seemed like there was a significant drop in  attendees – at least compared to years past or especially compared to Dreamforce.  Feels like there is a lack of excitement. Makes me feel like Oracle has hit it’s apogee.

  • Good News ? Oracle is not going anywhere.
  • Bad News ? Oracle is not going anywhere.

Oracle is solid awesome technology and is not going to disappear at the same time they’ve missed the cloud opportunity and not going anywhere in the industry. While the industry is moving to the cloud, Oracle has been concentrating on private data center mega machines (Exadata).  Now having Exadata engineered for Oracle makes it compellingly turn key and screaming fast, but why buy Exadata when you can buy push button power by the drip for a few dollars at time?  Who wants some aging hardware carcasses lying around as the industry advances? Even the Oracle buildings look like the old carcasses of hard drives . The future is clearly cloud and Oracle has missed the boat. Until about 4 weeks ago, you couldn’t even create a regular Oracle database in the Oracle cloud. From talking to people I respect and who work with Oracle cloud, the only thing Oracle’s cloud is really good for is Fusion, which works well in Oracle’s cloud (and is also too complicated to run in house). You may ask “What about all the massive Oracle cloud sales?”  Well, according to a number of articles out there Oracle is forcing people to buy these licenses which they may or may not ever use.

Oracle as a long history of denigrating competing technology X only to turn around a few years down the road after developing technology X in house and saying “we’ve got technology X, we are the best at technology X and you can’t live without us and technology X”. It’s worked in the past but this time it’s failing.

One reason it’s failing is because Oracle doesn’t seem to get cloud. I heard a number of Oracle people at OOW laugh and chuckle that Cloud is just time sharing.  Cloud isn’t about time sharing it’s about fast, agile, power. Cloud is about power at one finger tips with out hiring a team of experts and with out buying tons of hardware. Instead, with cloud, we can pull out our credit card, spin up as much power as we want and shut it all down when we are finished. Cloud is about speed. Cloud is about having the best experts in the industry design and maintain systems for us. Think about it. For example … are you worried about security? How much do you spend on security? How much does Amazon spend on security?

I once heard that Larry said, after buying Sun, that he didn’t need innovation in house. He could just buy innovation. Maybe. Who knows what tricks Oracle has up their sleeves. Larry has a long solid history of amazing business success and strategy. When the rumors started that Oracle might buy Salesforce, I thought, “wow that would be awesome and put Oracle straight in the middle of the cloud revolution and maybe Benioff could be a successor to Larry.” but alas those rumors were reported to be wrong.

Larry famously called Google “evil”  a couple of years ago. Given that he wasn’t giving many interviews in the preceding year I found this a shocking accusation. I ponder a long time as to what Larry’s strategy in saying this was.  My conclusion is that he realized for the first time that Oracle was seriously threaten by Google. Google was showing for the first time that it had the power and technical ability to challenge Oracle. Seems though in retrospect that the real challenger is Amazon.


Oracle cloud infrastructure play caps a month of big moves in cloud

Amazon To Build $600M Private Cloud For CIA




  1. Trackbacks

  2. No trackbacks yet.

  2. | #1

    Hi Kyle,

    > While the industry is moving to the cloud …

    I still doubt this (but maybe we Europeans are a little bit more skeptical in general) at large scale. The (producing) industry wants downtimes when they can live with it (e.g. shift changeover, pre-production or whatever). This is not possible with the (Oracle) cloud as the provider has a large shared environment and can not consider every customer requirement. The downtimes are dictated by the provider and not the customer. The end result looks something like this in consequence:

    … i have seen a lot of MSPs and even when you have dedicated environment, it is so hard to coordinate downtimes between internals and externals. Can’t imagine this in the cloud. However in my opinion the cloud is amazing for PoCs, playground or non-critical prod systems, but Exadata has a different target group.


  3. Noons
    | #2

    Agree with most of this. I’ve made the very public comment many times in the last 10 years that Oracle were shooting themselves in the foot. Big time!

    It all started with the demented attitude that only “community recognized experts” (or is it “evangelists” now?) were worth of hiring. Great for those who made a living out of selling “hit-and-run” consultancy. But fundamentally flawed in the medium and long term…

    Oracle completely ignored that average, run-of-the-mill DBAs were the very people that promoted their products over all others, since their history began.
    Result now? In many places you can’t find a single DBA that will recommend ANY Oracle products.

    On top of that, Oracle completely turned their ear away from the many, many complaints and suggestions of these same average DBAs about the stability and security of their products and the extreme hardship of installing them and making sure they worked reliably. Let’s not even mention the constant patching and the disruption and extra costs that causes.

    Result? They now get attendance replies of <6 (yes, less than six!) to their conferences and seminars here in Australia!

    Of course the fault of all that is with the DBAs, not Oracle! And "shoot the messenger" remains the very successful tactic of incompetent marketeers and managers…

    Then they bet the farm on java and (con)Fusion. Which as expected turned out a disaster for many years. I won't go into the details of why, done so many times elsewhere.

    Then they decided to bet everything on their own (Sun's?…) hardware and for a long time ran in it only Oracle products (Exadata, ODA, etcetc).
    Completely ignoring that IT data centres don't just run Oracle products.
    A modern IT department runs a gamut of many, many databases and applications in their data centre(s), of which Oracle products are just a (minor) part.

    Interesting also why they chose to make Sun such a preponderant part of their make-up: for many decades Sun was well known for pushing only competing products to Oracle – Sybase, Ingres, etcetc. I can see all those "defenders of the faith" now willingly working like mad to make Oracle a success (NOT!). Talk about sap(!) work…

    I can't even begin to imagine what all that is doing to their "cash cow": the maintenance and support license income…

    As a result of pointing out the above very publicly and very often, I lost count of the number of times Oracle sales reps and "partners" called my management offering freebie "competent consultancy and expertise" if only they got rid of me. Always good for a chuckle! :)

    But enough of Oracle's demented policies, marketing and bully tactics for now.

    There is one bit of your post that I completely disagree with. Where you mention the "advantage" of the cloud being a "turn on, run, turn off, pay only for that use", that being "the future" of IT.

    It completely ignores the simple fact that in-house IT is NOT used mainly for that type of computing! In-house IT is used for systems that are PERMANENTLY on, and need to be available and with data current and consistent, for years. Nay, decades!

    Development and test stuff? Almost in total agreement.
    Real production? As in "if it's not there, you might as well start writing your resume for a new job"? Not in your life!

    And the notion that the only place to find IT expertise is in the cloud is soooooo wrong, I won't even touch it! :)

    But of course: Oracle apologists will throw heaps of cloud at all the above! Hey, I started a private cloud for databases (Oracle and MSSQL) in our data centre in 2008, long before the term was even in use anywhere else. Does anyone really imagine I don't know exactly and precisely, first hand and in practical sense, what I'm talking about?

  4. Narendra
    | #3

    Good post Kyle.
    Couldn’t agree more with you, especially on Fusion. BTW, I thought calling fusion a “con” (hence con-fusion) was my invention. Glad to see (at least) some agreeing the over-hyped Fusion. I will still stick to my opinion that Oracle only produced one good quality piece of software which is RDBMS…rest all are…

  5. khailey
    | #4

    Just today this article came out on the growth of AWS.

    From inside a data center, especially at some old, crusty IT department of some large corporations, it may seem like any significant usage of the cloud could never happen, but looking at the growth rate, it’s hard to say the impact won’t be huge.

  6. Noons
    | #5

    Not at all surprised. AWS got it all sorted out in a very compelling way. To a similar extent, so did Google with their “Big Query”.

    Fact is: we’re using AWS already and have been using cloud for nearly 2 years.
    The plan is for a LOT more use of AWS – for our “big data” stuff.
    Around 200TB of it – mostly live real-time.
    And we fired no one to get there – but we hired a “data scientist”!
    Not bad for an old crusty IT department of a large corporation. 😉

    But we also have to run our payroll, billing, our data centres in each of our near 60 shopping centres, plus corporate accounting, reporting, planning, etcetc.
    AWS or any other cloud helps us in that by exactly 0 (zero) as they are not transient tasks or on-demand and are perfectly cheap the way we run them now.

    Let’s not forget that “cloud” is not a replacement for applications and their execution in a computer. It’s simply a provision of a facility to do those. On demand or permanent, depending how much one wants to pay. There is no “new” anything in the cloud in terms of IT processing. What there is with cloud is a new way to execute (and pay for) that IT processing. Computing facilities in the cloud are exactly the same as outside, it’s not a “new model of computer”! Prior knowledge of computing doesn’t vanish because one uses the cloud!

    It’d cost us more to investigate alternatives for everything than what we spend in-house. And let’s not talk about how expensive it’d be to integrate new solutions to old data in a total cloud environment. Yes, we use our older data for history analysis and future extrapolation – doesn’t every corporation with more than one brain cell?

    Although that doesn’t mean we won’t use more AWS cloud if/where we see an advantage.
    My estimate is that in 3 years time we’ll be 50% (in number of applications) in the cloud, with possibly an extension to 75% in < 5 years. Mostly IaaS and PaaS. But 100%? Nope, that just won't happen.

  7. khailey
    | #6

    yeah – that’s the confusion about Cloud. Most people think that Cloud is being pitch as some new type of computing, but that’s not it. It’s about ease and agility. It’s like

    Make it easy or die (software is eating the world)

  8. khailey
    | #7

    Or as Larry said himself at OOW

  9. | #8

    Very interesting, but I think you are wrong about where Oracle is going. I agree that Exadata and M7 and so on will have a relatively small role to play in selling on-premise systems. I believe that they will, however, be the path through which Oracle offers the best performance for what will be a competitive or perhaps even close-to-lowest price for cloud computing. I can also tell you that the dominant viewpoint within Oracle up through many chains of management is vastly different from “Oracle people at OOW laugh and chuckle that Cloud is just time sharing”. Are we relatively late to the game? Yes. Are we going to catch up? Good question. In the past Oracle has “caught up” just fine – now everything is moving so fast. Except enterprise -> Cloud. It’s too soon to tell how it will all play out, but I would say your assessment is overly negative and out of synch with the current state of affairs inside those massive hard disk-like buildings.

  10. khailey
    | #9

    As much a I would like to see Oracle win, the evidence paints a less than bright forecast, though it’s not over till the fat lady sings : “Amazon, Microsoft, and Google all reported bang-up earnings last quarter, not least because of their big bets on the cloud. ”

    Oracle also says its cloud business is growing, but there’s a catch: Oracle is going to its existing customers — which is just about every large company — and using strong-arm tactics and discounts to get them to try Oracle’s cloud.

    The current Oracle Cloud is fine for existing Oracle customers, letting them run their Oracle databases and CRM apps at larger scales. But cutting-edge adopters who would be more willing to jump to the cloud are likely to be leaving Oracle anyway. And those left behind are skeptical of this whole “cloud” thing, unwilling or unable to try something new.

    “Oracle knows it can’t reach broader developer markets,” Bartoletti says.

    That trend is great for end users, who get cheaper, more flexible infrastructures. Similarly, it’s great for Amazon and Google, who have no legacy enterprise businesses to protect. It’s “all upside,” says Anderson.

    But “if you’re an Oracle…yikes,” he says.

    Just know it won’t last: The allure of hyperscale in the era of software eating the world is just too powerful.

    “All of them will go to Amazon or Microsoft in the end,” Anderson says.

  11. | #10

    Anyone remeber Mainframe? So yes, it has all been there including the NSA backdoors. We moved away from Mainframe to Client/Server technology, because of the vendor lock in and inflexible stacks. For that we traded virtualization, dynamic compute power, highly specialized Mainframe Admins against cheaper unstable number crunchers and poor I/O subsystems. Now that we finally have these poor design up to the level Mainframe already had 20 years ago, ending our Oddysey, we get lured into the “Cloud” a bill of material commodity hardware with an automation stack, cos that is what all these Engineered and Cloud Systems are at the end. So the question here is: do you wanna play the same game with low entry level cost, but slow vendor lock in again or save the money / keep the tactical advantage of having your own it or not? Do you really want your car design drawings and chat discussing stored public and analyzed by NSA and Chinese government? Does you company live from innovation or the stock trading hype created by your CFO and Marketing guys. If it lives from the hype, don’t worry, go to the cloud: you have nothing to loose 😛

  12. khailey
  13. khailey
    | #12

    @Efstathios : what we are seeing across 100s of Fortune 2000 is that adoption is faster in US west coast companies which is not surprising as Silicon Valley has the mentality of change and innovation. If we look at the US heartland, uptake is much slower to non-existant. One example is when I give talks on the West Coast, most people have heard of Docker and DevOps. In the middle US or outside of the US, sometimes no one in the audience has heard of either. Europe tend to be even less cowboy than the US in general so it’s not surprising you are not seeing uptake … yet

  14. khailey
    | #13

    By 2018, at least half of IT spending will be cloud-based, reaching 60% of all IT infrastructure and 60% to 70% of all software, services and technology spending by 2020.

  15. Peter
  16. khailey
    | #15

    another post , this time from Curt Monash, on the decline of Oracle

  17. khailey
    | #16

    Another article detailing large companies going all in on AWS like Juniper, Netflix and Intuit and that hardware giants like HP, Dell and even Oracle should be terrified:

    I think Larry knows it’s too late. I think he first started realizing it when he called Google evil

    After this interview I was like “why did Larry call Google evil” and I think it was because for the first time he saw the writing on the wall. These young upstarts were actually going to push Oracle off the main stage. Nothing more frightening than that I imagine if you are Larry.

    Now it might not be Google as much as Amazon but the context is the same.

  18. Noons
    | #17

    Larry should have done one thing 10 years ago: retire.
    It’s obvious he’s been out of his depth since then.
    Worst thing he did was to bet the house on Java + 3rd world developers and the only computer company that hated every Oracle account on sight: Sun.
    I lost count of the number of times I had insulting remarks made to me by Sun officials every time I installed Oracle in one of their systems for a demo: it was obvious they hated the guts of anyone who could spell the name.
    Same for HP. What does he go and do? Gets in bed with the guy at HP who caused the problems way back then.
    Hey, it’s history and fact. Not “prediction”. I leave that last one to the “experts” in the industry. Meanwhile, I just get on with the job, like I have done for the last 40 years. I must be doing something wrong…

  19. khailey
    | #18

    More data, direct from Oracle
    “On-premise software and hardware sales fell, but cloud revenue surged with the exception of infrastructure-as-a-service.”
    clearly Exadata is a loosing market and cloud is the the now and growing fast and Oracle is way behind when it comes to cloud

    ‘see also

  20. khailey
    | #19

    another one
    “demand is shifting away from its database, middleware, and hardware businesses as enterprises move workloads out of corporate data centers and into cloud infrastructure vendors like Amazon Web Services (AWS) and Microsoft Azure.”

  21. Noons
    | #20

    I really don’t think “enterprises move workloads out of corporate DCs”.
    What is happening is folks got totally fed-up with the complete lack of any QA on Oracle 12c.
    Which even now – years after it was declared “ready” – needs a truckload of patches before it can be installed or used to upgrade prior releases!
    And as a result, they are moving in droves AWAY from that database. Not necessarily into the cloud: AWAY from Oracle!
    They might not have access to the usual RAC/ODA/Exa* nonsense as a result, but few folks need those anyways!
    Looking at our landscape, we are now (10 years later) on over 400 MSSQL dbs with less than a dozen in the cloud. Oracle? We still have only 30, including all development/UAT ones, none in the cloud.
    Cloud is not an insurance that Oracle databases will be less buggy or need less patching.
    What we need are stable solutions that work.
    Where they live is immaterial!
    This is the thing that no one at Oracle is listening to: RDBMS 12c is as unstable as an egg on top of a nail head and NO ONE is doing ANYTHING to improve the situation!
    This includes those who are supposed to be the “evangelists” and “advocates” of the product!
    How’s about LISTENING to complaints and doing something instead of belittling the messengers or plain ignoring them?
    Hey, we are all stupid and will continue to fork out loads of cash for what is essentially a low-value proposition?
    Yeah! Right…

  22. khailey
    | #21

    Yes, seeing more and more momentum to move off of Oracle.
    Though is unclear to me that the pain (money, time, resources) of moving to a new database are out weighed by the benefits.

two × = 8