Still NOT Positioned For Analytics
Oracle’s Big Data Appliance, originally announced in October 2011, is now officially for sale and includes the Cloudera distribution of Hadoop. Despite the inclusion of Cloudera, Oracle's position on Hadoop remains fundamentally different than IBM. IBM embraces Hadoop for all use cases, especially as a platform for analytics. Oracle continues to position Hadoop and their Big Data Appliance strictly as a platform to acquire and transform unstructured data to be loaded into Oracle database for analysis.
“Oracle Big Data Appliance is an engineered system optimized for acquiring, organizing, and loading unstructured data into Oracle Database 11g”
A Hadoop appliance should include more than just pre-integrated components from the Hadoop ecosystem. It should have the elements that enterprise clients need to implement and execute a big data ANALYTICS strategy. It also needs to tie into to a complete big data platform view that includes a mature Hadoop distribution, a solution for streaming analytics, and an MPP Database.
Furthermore, the point of an appliance is to speed delivery and make things simple. With this appliance all you are getting is a Hadoop distribution shipped on a box that doesn’t include the tools, accelerators, etc necessary to get the customer up and running quickly.
What about Cloudera
In order to capitalize on the buzz around big data, Oracle has been forced to partner with Cloudera to include Cloudera’s CDH3 Hadoop distribution in the Oracle Big Data Appliance. This effectively means that Oracle’s Big Data Appliance is a very large, expensive, Cloudera system. Oracle joins Dell, SGI, and Netapp in reselling Cloudera on their respective hardware platforms. Cloudera customers now have a broader choice of bundled hardware solutions, but Oracle’s has got to be the most expensive.
Cloudera will provide level 2 and level 3 support while Oracle will be taking all of the initial support calls. Customers calling for support will likely be bounced around. It seems like Oracle perhaps doesn't have Big Data expertise in-house.
Oracle is also now dependent on Cloudera for what is arguably the key software component of the Oracle Big Data Appliance. IBM has no such dependency.
Starter System “only” $450,000 plus optional $216,000 connector software
The price tag for Oracle’s Big Data Appliance is $450,000 (plus 12% annual maintenance). Add in the optional Oracle Big Data Connectors, and the price jumps another $216,000 (plus 22% annual maintenance) to $666,000. Considering that most customers are just starting out with Hadoop, this is a very steep entry price.
Unlike Oracle, IBM offers customers a variety of ways to get started with Hadoop for a much lower cost with InfoSphere BigInsights available in the Cloud via RightScale, directly on Amazon, Rackspace, or IBM Smart Enterprise Cloud. IBM also offers cost-effective, incrementally scalable, System x hardware solutions for BigInsights deployment in customer data centers.
For those customers who do have a need for a large Hadoop system, IBM’s systems offer flexibility and cost advantages. “Curt Monash, president of Monash Research, says organizations can set up big data infrastructures that are just as good and much less expensive than what Oracle is offering in its Big Data Appliance. He said he isn’t really buying the integration point that Oracle is selling. There are downsides to a big data appliance, including “overpaying for the same thing you could get much more cheaply elsewhere,” and “losing out on some flexibility because you’re stuck in a fixed appliance format,” he said.
What are the Oracle Connectors?
- Oracle is bundling the following software components and selling them together as a $216,000 option for the Big Data Appliance.
- Oracle Loader for Hadoop
- Oracle Direct Connector for Hadoop Distributed File System (HDFS)
- Oracle Data Integrator Application Adapter for Hadoop
- Oracle R Connector for Hadoop
These components enable the Oracle Big Data Appliance to interoperate with the Oracle Database and drive up the cost of the Oracle solution.
- Oracle Big Data Appliance is not positioned for analytics
- Oracle says it is intended to acquire and transform unstructured data for loading into Oracle
- The appliance starts at $450,000 plus $216,000 for optional Oracle Connectors
- It is not suitable for customers just starting out with Hadoop – too big and expensive
- Includes Cloudera, but Cloudera customers can get alternative hardware solutions cheaper
- IBM BigInsights is available as software only, in the Cloud, or on IBM System x reference architectures – all of which offer more flexibility and lower cost than Oracle