Two weeks ago, HP and Microsoft announced they would jointly spend $250 million to better integrate their hardware and software systems. That's one of the first reactions to Oracle's expected acquisition of Sun Microsystems and its entry into the hardware business. More such alliances are likely to spring out of the ranks of Oracle's competitors.How do you compete with a potential powerhouse combination of two industry leaders? On the one hand, you try to take away part of the low end market, where Microsoft thrives, with an open source play. Oracle learned this trick from watching IBM. IBM respects open source as a broad social and industry development, and uses it when it suits its business purposes. Far from killing it off, Oracle will try to do the same with MySQL.
With the acquisition of Sun Microsystems, Oracle finally has a weapon with which to pummel Microsoft, as Microsoft has been pummeling Oracle on the low end of the database market. Oracle will use MySQL the same way that IBM used Linux -- to cut the ground out from under a competitor who might otherwise show persistent strength at the low end of the market. Linux ate into Sun's Solaris revenues, and we all know how that story's going to end. Oracle has limited interest in selling the Oracle database system to the small development teams and independent developers who use MySQL. The Oracle system is too big, too complicated and requires professional database skills to administer. On the other hand, it isn't interested in seeing its most viable software competitor build more strength from the low end of the database market.
MySQL has already run off with a large share of the low end, particularly Web application developers, and Oracle would like to see it continue to do so. It is likely to make MySQL stronger in the ways that compete with SQL Server but not so strong that it competes with Oracle itself. MySQL is simple to use and implement in applications that require rapid, read-only access to data and content, much simpler than its open source competition, PostgreSQL and Ingres.
MySQL had an excellent team behind it in Michael "Monty" Widenius, David Axmark and former CEO Marten Micklos. Development talent was harnessed to organizational talent, and both were capable of communicating direction along with credibility that the system was going to be developed rapidly.
That was then, and this is now. The team has dispersed. MySQL is about to go into the Oracle organization, where it will thrive just as long as it is useful to Oracle's long term strategic direction. And while MySQL has its strengths, it also has its weaknesses in lacking its own storage engine. You select the storage engine you want, then plug it into the database engine. Two of the most popular for use with MySQL, InnoDB and Berkeley DB, are already owned by Oracle. If you use MySQL in your business, it doesn't matter that you got it for free. Chances are, you are in hock to Oracle already and will end up, one way or another, paying Oracle for support. That, plus the fact that MySQL probably displaces a potential Microsoft SQL Server sale is what they call a win-win situation at Oracle. When HP finds a new hardware competitor on the block, it tends to look around for new partners, de-emphasizing any partnership with the interloper. It's already done that with Cisco Systems. If HP more tightly integrates its data center management software with Windows Server, including the Hyper-V hypervisor and Microsoft's Systems Center software, that will give HP an advantage in selling x86 servers that the Oracle/Sun combination will find hard to match. Another way you compete with the new powerhouse is kick the level of competition up a notch, to the cloud, where Oracle and Sun may have a harder time matching what you can do. HP and Microsoft together are more likely to enable "private cloud" construction in the enterprise data center than either one is alone. At the same time, special efforts can be made to coordinate the internal cloud with what Microsoft will be offering through Azure. In the future, much software development is going to take place in the cloud, using the tools that Microsoft and other parties will make available there.
Again, if Oracle wants to compete in the hardware field, it seems to have missed the fact that it will need some long term strategy that allows its hardware to function well in the data center and in the cloud, with software coordinating operations between the two. What is Oracle's plan in that regard as a new hardware player? I can see how Microsoft and HP are going to have such a plan; the competition just got a little stiffer for this new combination, Oracle and Sun.
InformationWeek has published a new report on cloud governance, risk, and compliance Download the report now (registration required).