Oracle-Sun deal: Integration issues?

Through one important piece of corporate computing jargon - 'integration' - Oracle has found a justification for its $7.4bn acquisition of Sun.

Now it will have to convince sceptical customers, too, that the idea makes sense.

The all-cash acquisition agreement - announced on Monday, costing Oracle $5.6bn with Sun's cash factored in, and expected to close this summer - puts the innovative but financially bumbling Sun out of its misery after IBM's move to buy it fell apart earlier in April. The way to fit Sun's technology into Oracle's business model goes back to a project called Raw Iron that is more than a decade old.

Raw Iron ideas placed application software front and centre while demoting the server hardware itself and the operating system to a subordinate role. The customer who needs some database software hardly needs know what's going on under the covers.

What is smart about the approach is that it lets Oracle profit from Sun's diverse technology - which includes not just servers but also open-source software, including Java and the MySQL database that Oracle already tried to buy years ago - without disrupting its own business too much.

Oracle signed a Raw Iron partnership with Dell and worked on it with Sun, IBM and then-independent Compaq. With Sun's technology in-house, one major challenge of those deals - that is, who is in the driver's seat - evaporates with Sun being a part of Oracle. There is no longer any question about which partner owns the customer relationship, which services the technical support contracts, and how the sales revenue is divvied up.

But Raw Iron, along with the related concept of server appliances that arrived a few years later, was a marketplace dud.

Customers appreciate integrated technology to an extent, but Raw Iron and server appliances quietly submerged beneath the waves. Also worrying for Oracle is the failure of one of its integration ideas, Unbreakable Linux. Customers by and large ignored this Oracle attempt to offer its own version of Linux, a clone of market-leading Red Hat's product.

Oracle chief executive Larry Ellison is a true believer, though, making the sales pitch in the company's official statement: "Oracle will be the only company that can engineer an integrated system - applications to disk - where all the pieces fit and work together so customers do not have to do it themselves," Ellison said. "Our customers benefit as their systems integration costs go down while system performance, reliability, and security go up."

He has a point. Sun has always focused centrally on the database market, and it has compelling technology assets for it that it has been unable to sell effectively: its current Niagara and the delayed higher-end Rock multicore processors, its Solaris operating system, and its Thumper storage servers with tremendous built-in data capacity.

And selling products at this high level of integration gives Oracle a way to ingest Sun's considerable open-source assets - among them Java, MySQL, Solaris, GlassFish, NetBeans - without too much indigestion. It might even give Oracle some incentive to be more active with the open-source community that it has kept mostly at arm's length.

Another issue, however, is that server appliances are to an extent an artefact from an earlier era, when companies bought and managed discrete systems. That remains a big business, but it is at odds with two important trends gaining steam in the industry.

First is virtualisation, chiefly through EMC's VMware software. This lets a single server run multiple operating systems, with the software collection moving flexibly from one physical machine to another as work load demands shifted. By breaking the hard link between hardware and software, virtualisation undermines the integration sales pitch and inserts a third party's technology between the server and its higher-level software.

Second is…

For page two of this article click here