Oracle on path to become the Apple of the enterprise

Oracle on path to become the Apple of the enterprise

Summary: With its plans for a proprietary cloud along with its encouragement of customers to use its software on its engineered systems, the database giant is looking more and more like it wants to become the Apple of the enterprise.

TOPICS: Cloud, Apple, Oracle

Oracle looks set on becoming the Apple of the enterprise.

At Oracle OpenWorld in San Francisco this week the company has used every keynote speech and press briefing to put across one message: with the advent of a fleshed-out cloud, Oracle is in a position to make almost every hardware and software component that an enterprise may need. No other enterprise company can do this, Oracle says. 

Oracle seems to be trying to become the Apple of the enterprise. (Composite image: Jack Clark)

"We have service where we [can] patch the entire thing from the database all the way to the firmware in the controllers," John Fowler, Oracle's executive vice president of systems, said in a keynote speech on Wednesday.

The closest company to Oracle in terms of this approach is not one of its standard rivals - SAP, HP, IBM - but Apple.

Oracle is aiming to be able to sell a customer the hardware on which they perform computing and storage, the software to run jobs onto it, and a raft of cloud services which they can store their data in and use to update their software. 

To me this approach is the same as Apple's, which designs its own hardware, writes the OS which its systems use, and, via the App store, controls a widespread developer ecosystem.

The similarities go further when you look at what both companies do with their respective component ecosystems.

Like Apple, the majority of vendors that Oracle procures components from are bound by non-disclosure agreements, preventing them from crowing about having kit in the latest engineered system. In the same way that companies like LSI and Fusion-io are not able to confirm whether their technology is used within Oracle's systems, Apple component suppliers cannot officially claim to be inside an iPhone or a Mac computer. (There are exceptions, such as Intel, but they are few and far between.)

Exadata leads the way

The strategy comes across most clearly in the updated range of Exadata systems. With these bits of kit, Oracle have "been able to optimize all the pieces to work together – much like an iPhone is an 'engineered system'," Scott Swigart, co-founder of Cascade Insights, a market intelligence firm said.

"Here the argument is about price per performance. Oracle claims that for X dollars you could buy a bunch of HP gear, and configured it yourself, but you wouldn't get the same performance as if spent the same X and just bought an Exadata."

Though sales have been disappointing for its hardware, the customers that do use it seem to like it. 

"The first time I heard about this engineering of hardware and software together I thought that sounds like a good idea. Having seen it in reality and lived with it now I can see it does pay dividends," Marc Terry, managing director of commercial services for VocaLink, said.

VocaLink uses Exadata systems to help it process transactions between tens of thousands of UK cash machines and banks, along with handling the BACS system which is used to pay the majority of wages in the United Kingdom. He described the hardware as being both "cost-effective" and an "exceptionally high-performance system."

When asked about worries over lock-in, he indicated that if Oracle maintains a combination of good service and strong hardware, he would remain a customer.

"If the technology is reliable and it doesn't break and we get performance because it's engineered together and that outweighs the potential from having multiple sources, then we'd go with that," he said.

From Oracle's perspective, I imagine the motivation to create a 'red-stack' is that this gives the company the greatest degree of control over not only its products, but its customers as well.

Once everyone is on a standard platform that you own, then driving major changes becomes a lot easier. In the same way Apple chucked out Google Maps in the latest iOS and replaced them with its own software, Oracle is able to remotely patch its systems to suit its future purposes.

A risky strategy

However, history is littered with enterprises that have failed to handle both software and hardware. Today HP's CEO Meg Whitman told analysts that because she had inherited a bloated company, profits would be down next year and it could be years until HP's various business divisions recovered. In the mid-2000s IBM, conscious of the types of problems that have subsequently befallen HP, got out of one of its hardware lines by selling its personal computer business to Lenovo. Similarly, Dell tried briefly to get into tablets, failed, and has regrouped around core services.

Ultimately, for Oracle to become the enterprise equivalent of Apple it will need to focus around a few key products and relentlessly iterate on the design of them to please customers, otherwise it risks over-extension. Even Apple has fallen foul of this – at the end of 2010 the company got out of its minor 'Xserve' server business.

Amazon may be 'the iPhone of the cloud', but Oracle wants to be the Apple of the enterprise. Whether it can tie everything together and still impress customers remains to be seen, but judging by the announcements we have heard this week, it is committed to this strategy.

Topics: Cloud, Apple, Oracle

Jack Clark

About Jack Clark

Currently a reporter for ZDNet UK, I previously worked as a technology researcher and reporter for a London-based news agency.

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.


Log in or register to join the discussion
  • That's Nice

    But really, Oracle are more like the RIM in the Enterprise than the Apple. They hardly have the kind of market share that Apple enjoy. And last I heard no Oracle product was exactly user friendly.
    • RIM is better than that!

      At least RIM has a future.
  • Important point missing

    If you're writing about what they would like to achieve or become, it would be good to mention also what are the chances that they will succeed in that. Especially in a situation when those chances are close to zero. These guys behave like they would wake up from long sleep and think we are in the 80's. They proposals have not much in common with how IT looks like nowadays and what are the trends, even if they happily adopted fashionable words like 'cloud' into their marketing slides - the truth is that real offer below is the same as 20 years ago. They want to sell you mainframe and then take money from you for next couple of years. They will live happily still for long, taking money from many huge already locked-in corporations, but chances for their progress are zero if they won't change their attitude to making business permanently. And this is unlikely to happen with Ellison at steering wheel.
    Mr Wrong
    • Every chance of Oracle making good business in the cloud

      Given that there are many still unsolved problems with distributed computing, the share-nothing approach of so-called big data solutions is doomed to failure.

      RAC (or equivalent offerings from IBM or Sybase) are the only viable options for reliable cloud operation for real systems that do serious work in transactional systems.
  • Except that....

    ... the heart and soul of Oracle, which is their DBMS engine, runs on all kinds of hardware/OS configurations. From Windows server, to Linux (many distros) to UNIX (AIX,HP-UX and of course Solaris), heck even on OSX server.
    If they start pulling the plug on those, it will be their undoing.
    • if they start?

      They started already, first by trying to remove HPUX support, then by numerous hints that the only "safe" linux variant is their own. The rest is even a matter of time
      Mr Wrong
    • Then switch to DB2, SQL Server, Postgres, Sybase

      SQL-DBMSs are all logically equivalent to each other.
      • Wrong...

        Sorry to say but you couldn't be more wrong than that. Moving legacy business application to another DBMS is most probably the most complex and time consuming of all kinds of IT projects. That's why Oracle is so safe with their base revenues, most of their customers don't want to spend money and take a risk, they prefer to keep paying oracle and sleep well. Same like IBM with mainframes.
        Mr Wrong
        • Actually there are some far more difficult tasks

          Moving all the applications from say C# to Java or the other way round would be an order or magnitude more difficult than changing DBMS.

          Moving everything from SAP to Oracle Applications would be a lot more difficult than switching DBMS.

          Of course massive investment in the priorietory, procedural parts of Oracle like PL/SQL might hamper you quite a bit, but you could look on the migration as a good reason to get rid of all the old-fashioned procedural stuff.
        • My statement still stands

          All RDBMSs are logically equivalent to each other because every one is based on the relational model.

          However you could convincingly argue that a SQL-DBMS is not a relational DBMS. A point I would be forced to accept.

          Even so constructing your database model at the relational level should make migration relatively easy, compared to migrating most other technologies.
          • You must be a Newbie...

            You don't know what you are talking about. There are inter dependencies, different reporting tools, not to mention hundreds of users and their own custom written code...
          • I'm not claiming it's easy

            Just not as hard as some people make out.

            30 years experience in IT.

            20 years experience with SQL-DBMSs.

            How much longer do I have to work to stop being a "newbie"?
          • Mind you

            Most Oracle experts I have talked to claim that migrating DBMS is incredibly difficult, unless of course you happen to be migrating to Oracle...

            However, very few of them seem to be prepared to accept that Oracle is NOT relational (but mind you the competition isn't either).
          • How do you explain then

            How Sybase lost so much market share to SQL Server and Oracle?

            Obviously plenty of people have achieved this incredibly difficult task.

            They could all move away from Oracle too if the will was there.

            Shifting DBMS is hard work, but a lot easier than a lot of other migration tasks.
          • apples vs oranges

            Maybe you work with databases for 20 years (my humble person only 13), but it seems like you haven't seen much during that time. You mess up everything. It's relatively easy to migrate from simple dbms to complicated, like, there's not a problem to move app from mysql to any other db. Sybase and mssql are practically the same (at least they were exactly the same 10 years ago, now not anymore) so it's also no big deal. But it's damn complicated to move application from oracle to any other dbms. Or from db2/sybase to oracle. You keep talking about tables, data and relational model. Nobody said it's a problem to move data. Problem is with application code that practically needs to be rewritten, especially if as usual with oracle it is in 99% written in pl/sql. Moving data between different dbmses is piece of cake. Also it's not a big deal to migrate app that has 10 concurrent users, as then you can practically forget about completely different locking strategies, concurrency models etc in for instance db2 and oracle. But if you're trying to move app that has 2000 concurrent users, believe me, it is heavy task, close to being undoable.
            Mr Wrong
          • Plenty of big Sybase sites moved to Oracle

            I can think of one very large Telecoms company that moved its whole billing system from Sybase to Oracle.

            Perhaps you can explain why going in the other direction should be so difficult?

            I accept the PL/SQL argument to some extent, but with recent advances in SQL and optimizers there is less and less reason to use PL/SQL anymore. Many companies use PL/SQL in a procedural way which is wholly inappropriate to working with an RDBMS, I've seen code written with cursors in PL/SQL that could be replaced with a single SQL statement, and afterwards ran in 100th of the time. So even if you use Oracle it's probably time to start trimming the PL/SQL.

            Of course the applications, if they are well behaved, shouldn't be troubled by switching DBMS. I would agree that the main problem will be finding all the dependencies and doing the syntactic replacement of DBMS specific functions with the appropriate functions in the target database.

            From my experience many people who have worked exclusively with Oracle think that Oracle is somehow special and different. It's not, it's just another SQL-DBMS. If it was so special and different this would be a very good reason not to use Oracle - as clearly it wouldn't be an RDBMS anymore.

            I don't know much about DB2 locking and concurrency, but I would guess it doesn't use snapshot isolation, so writing the constraints that have to be written in code or triggers should be far easier in DB2 than in Oracle.

            I'm not denying that switching DBMS is a big project, but it certainly isn't undoable, even for big systems, and it's certainly an easier task than switching from SAP to Oracle Applications.
    • Not really

      Oracle no more supports the OSX Server. At least, not when it comes to the database itself.

      If you want to run Oracle on Mac, run it on a virtual computer running Oracle Linux.

      But of course various Oracle application run on Mac, they just have to connect to an Oracle database running on another computer.
  • poor analogy

    Apple, despite its best efforts over the years to be completely proprietary isn't. It's simply not realistic if a company wishes to survive more than a few years. We own several iThings and Macs. We source all of our upgrades, from drives and memory sticks to replacement batteries and LCDs, from non-Apple sources. Most of our apps aren't from Apple. And we multi-boot into non-Apple OSes. We use iTunes minimally and we don't rely upon iCloud (we prefer other Cloud resources which already contain many of our assets).

    Yes, we buy the basic hardware and OS from Apple, but from that point onward our relationship with Apple is virtually nonexistent until and unless we decide to purchase another iThing.

    Sure Oracle wants to own and sell the stack...wouldn't any IT vendor? In reality, a large majority of prospects might express interest in "one throat to choke" but that is not the way they buy IT. It's an all-risk no-leverage arrangement that scares the crap out of them.
  • Poor analogy

    The author's analogy to Apple is lost on me....couldn't the same be said of an IBM? They have their own hardware, proprietary OS (z/OS anyone?), software, databases, cloud offerings etc. etc. There's simply nothing unique that an Oracle brings to the enterprise that an IBM doesn't and I would submit that the reference to "Apple" is a lame one.
    Kia Ora IV
    • Yup

      Oracle just wants to be like IBM. They have a long way to go. Start investing in basic research too, Larry.
      Rabid Howler Monkey