Pumping it dry: $200 a barrel and $25,000 per CPU

When Oracle acquired BEA systems, I and others noted the significance of the loss of the only independent Java middleware vendor. With Oracle’s recent announcement of a price hike for their products, including WebLogic Server, this is no longer a theoretical issue. They have the oil, and they think they have existing customers over a barrel. The need for alternatives is now even more painfully clear.
In fairness, Oracle’s move is partly driven by the weakness of the US dollar, but the increases in WebLogic pricing are far greater than those affecting other products.
Some applications previously priced at $3,995 are now listed at $4,595 — up 13.1 percent — while database software prices increased 18.75 percent from $40,000 to $47,500 per CPU. Other prices increased approximately 15 percent, according to Wang's report. The price for BEA's WebLogic application server is now $25,000 per CPU, up 47.1 percent from its $17,000-per-CPU price prior to Oracle's $6.7 billion acquisition of the middleware software vendor in April.
This decision probably indicates two things: that Oracle justified the high cost of acquiring BEA (actually, over $8 bn) through its belief that it can make more money from BEA customers by raising prices; and Oracle’s expectation that, with no independent vendor left, there is not enough competition left in the Java EE application server market for customers to resist such a price hike. From the same article:
Some industry observers have worried that the acquisitions could give Oracle a near-monopoly in some markets. The Forrester report says the price increase for BEA WebLogic could reflect Oracle's dominant position in the application server market.
In a two-horse race in the legacy app server market between Oracle and IBM, both vendors might well take that view, effectively creating the OPEC of application server vendors. IBM Senior Vice President and Software Group General Manager Steve Mills recently commented that he is “not particularly concerned with competition" in this space, “particularly from open source offerings.�
Fortunately, for customers, however, Oracle’s assessment is probably wrong. It seems dangerous to hike the price of a Java EE application server at a time when it’s clear that the market is moving to alternatives to the bloated traditional Java EE application servers. Unlike the hydrocarbon economy, alternatives to poor and expensive solutions can be created in years, not decades. In fact, many WebLogic customers are already migrating to next-generation infrastructures. Now the TCO argument has also become 47% more compelling to migrate from a legacy platform sooner, rather than later.
Maybe this is the right move for Oracle financially. Sell it for $200 a barrel and make the most out of the last gasp of the WebLogic revenue machine.
Steve Cobbs says:
Added on June 25th, 2008 at 7:52 amRod, you've been predicting the death of expensive app servers since 2000, why would we believe you now?
Jenn says:
Added on June 25th, 2008 at 10:49 pmGiant markets don't collapse over night, but certainly closing the size of a market to a few over-priced vendors could be such a catalyst. Consider the affect of ISVs and consulting shops, the average IT firm that comes in and has to integrate 10 different apps on some platform. They, of course, want to charge as much as possible with as least cost to them. Do you think they want to pay for expensive licenses or pass that along to the customer?
The past few years have also had a very quick evolution of technology standards.. from struts to JSF, EJB1 to EJB3, from J2EE to JEE, from JAX-RPC to JAX-WS, introduction of SipServlets (not 1, but 2 specs), AOP, etc. Unless "the big 2" are going to offer free upgrades from WAS/WL to the latest to get the new features, there is a very real and compelling reason for developers to seek alternatives, for the very least to use the new technology.
Lars Tackmann says:
Added on June 26th, 2008 at 3:34 amThis had to happen, Jetty/Tomcat with Spring is so popular these days that WebLogic is going to see a further decline over the next years. GlassFish is the only other appserver I would use if I have the choice. BEA does have a strong hold in certain middle ware areas, like its tuxedo services (since its 2 phase commit actually works).
It is going to be very interesting to see what will happen in the app server over the next few years. I do not see the spring source server going anywhere since it currently does not offer programmers anything special - sure OSGi support is nice, but that is so marginal an edge that it is not going to move anyone.
Kristof Jozsa says:
Added on June 26th, 2008 at 5:28 amI think I just liked Spring much better before the marketing machine arrived. Honestly, I think that was the main discriminator compared to the rest.. There's no way back, right?
Jin Chun says:
Added on June 26th, 2008 at 8:24 amAt some point, the critical slope will be reached on the database side as well. This has been less "accessible" just because of the scope of potential contributions from the wild compared to anything java based, even an app server, still is more accessible than a database kernel. It will be interesting to see how things move. This is going to hit a lot of companies that are currently trying to trim down their IT expenditures and will undoubtedly provide more of an umph to look at alternatives like MySQL, Postgres, Sybase, and others and the support around them. Looks like the BEA pricing has just been launched into the Websphere. The point often missed is that the TCO between these and commercially supported open source, isn't just a factor of 4-5, its much, much, much greater than that. I'd rather hire a few more good developers than pay for the license for medium size box.