Something is seriously wrong in Larry Land. Oracle does not command absolute control like it once did. You can see this clearly with the earnings the company posted last week and the growth that startups like Datastax are witnessing as more customers seek alternative databases for online applications.
Until this past week, the extent of Oracle’s problems were not known. But there is a cut, a slight bleeding that’s now visible. But how deep is the cut? How much is Oracle really bleeding? That’s exactly the question analysts asked in a Reuters story after the earnings results:
“Data base revenue, which has been the cash machine of the company, has changed. There are now alternative databases, as well as the cloud,” said Mark Moerdler, an analyst at Bernstein Research. “That pressure is still a tiny bleed, but it is out there and the question is – is it bigger than we think it is?”
We know this much. Oracle reported this week that new software licenses are down two percent. And that decline is in part reflected by the adoption of NoSQL databases offered by Datastax and a variety of other services that use in-memory technology and new SQL offerings at the database layer. Update: Of course, there are competitive forces at play with enterprise giants such as SAP, IBM and Microsoft playing their part. But it’s the startups that represent the core of the innovation.
The reason for the drop has more to do with the enterprise acceptance of online applications more than anything else, said Datatastax CEO Billy Bosworth in an interview last week.
That’s the truth. NEA Ventures Scott Sandell said to me at SXSW that CIOs are convinced to move their workloads but cloud security is still an issue.
That’s where companies like Datastax enter the picture. Datastax is built on Cassandra, a high performance, fault tolerant Apache open-source database technology.
Here’s how Bosworth described it to me in February at the Strata Conference.
Datastax, founded in April 2010, finished its first year with 26 employees. It ended 2012 with 100 employees. Bosworth expects to have 160 people on staff by the end of this year.
Customer growth has increased significantly. By the end of 2011, Datstax had 27 customers. One year later it had 270, with 20 from the Fortune 100.
Several dozen of those customers have moved either all or parts of their application off relational technology such as what Oracle provides.
When companies come to Datastax, they say the number one thing they need is security, Bosworth said. They are building from day one to avoid disaster scenarios.
Datastax, like other NoSQL providers, spans its database technology in a fully distributed way, across private data centers and the cloud.
Datastax differentiates by offering high performance at scale but without complexity.
How customers use Cassandra reflects on why Oracle growth has begun to stall. Often, customers will continue to use Oracle databases but will put it deeper in the backend. They will take another piece of the app and put it on Datastax.
Customers will build in a middle layer of services components that allows the app to decide which database to use for which workload.
Lighter Oracle workloads means less revenues, which we see reflected in the company’s earnings.
To counter this swarming hive of distributed systems, Oracle has taken the opposite approach, building out engineered solutions with their software running on big, new age mainframes. That strategy does not seem to be working very well. Oracle bought Sun Microsystems with plans to sell the hardware with its software.
Analysts tend to agree:
“The problem is, the growth of SaaS (software as a service) applications is undermining that strategy. When you subscribe to salesforce.com, you don’t need to buy a database, middleware or hardware,” said Patrick Walravens, an analyst at JMP Securities in a Reuters story last week.
It is unclear how Oracle has fared with its $5.6 billion acquisition of Sun Microsystems. And there is little proof that companies are going to start using one company like Oracle for all their hardware and software needs. Instead, they will mix Oracle software on commodity systems. Or they may even go with the new open-source server technology coming out of Open Compute. They have plenty of other options, too. OpenStack, the open cloud effort, is growing fast, as is Cloudstack, the open-source cloud service now part of the Apache Foundation.
Datastax has its own challenges. It competes with Amazon Web Services and all the other NoSQL providers such as 10gen. The ecosystem is still quite young. Finding qualified people is a challenge. Developers need more education, a change in thinking for the new cloud approach.
But overall, it’s clear that Oracle really is starting to show the pains of being an aging innovator. The earnings show a slight cut. The question is how deep the cut is and how Oracle will respond to challengers like Datastax.
[Correction: I incorrectly reported in this post that Oracle has not posted a profit since its acquisition of Sun Microsystems. Further, the profitability of the Sun deal is still a question mark, though complete proof can only be determined with financial proof from Oracle, which does not break out such level of detail in its financial statements. I apologize for the error.]