Oracle always seems to be able to pull off a surprise and today announced the acquisition of privately held Nimbula. Nimbula has $20M of investment from Sequoia Capital and Accel Partners and was an interesting SA/SV (South Africa / Silicon Valley) based start up More interestingly it was founded by two former Amazon EC2 development honchos, Chris Pinkham and Willem van Biljon. Oracle did not even dedicate a press release to the topic - just a short statement.
![]()
Nimbula had a great founding team with Pinkham and Biljon, who applied some of their lessons from building the Amazon's EC2 cloud into their private cloud offering called Nimbula Director. With a great focus on the power of self service, the product offers one of the fastest and easiest ways to get a private cloud going on x86 raw iron. Kudos to the Nimbula development (and marketing teams to expose this) to adopt a persona based development of the product, catering to the roles of cloud administrators, tenant administrators and end users.
After a short attempt to liase with Hadoop and a partnership with MapR - Nimbula went to support Openstack, opening their Director product to Openstack APIs.
Oracle in comparison had no similar offering, anything for orchestration across the private and public cloud. The work ahead will be to let Nimbula Director work and support Oracle's virtualization products. Moreover, once the ACME Packets acquisition will have happened - the IP will improve firewall throughput siginifcantly, which is critical for the performance of private / public cloud offerings.
On the other side it does not hurt that Nimbula was a key RedHat partner (though RedHat put their chips down already, with the acquisition of ManageIQ, which left Nimbula in a somewhat weaker position. Likewise VMware was supported by nimbula - and VMware is closer as a competitor to Oracle given the focus of Oracle in the virtualization area.
![]()
I expect also less hyperbole around Oracle's False Cloud - as you would have to argue that an OpenStack compliant product isn't a cloud product. And that would make all OpenStack vendors False Cloud vendors. Ironically that's something that could happen, if the elephants at the table at the OpenStack board meetings play it out as it once happened to ... UNIX.
Nimbula's meets ... Oracle
Nimbula was founded in the hey day of private cloud euphoria that prevailed 2-3 years ago. Pretty much all private cloud vendors stumbled over the reaction of the entrenched enterprise players - who curtailed the success of the smaller startups with all actions at their disposal. And then enterprises remained conservative - and kept asking - can we just keep buying VMWare.Nimbula had a great founding team with Pinkham and Biljon, who applied some of their lessons from building the Amazon's EC2 cloud into their private cloud offering called Nimbula Director. With a great focus on the power of self service, the product offers one of the fastest and easiest ways to get a private cloud going on x86 raw iron. Kudos to the Nimbula development (and marketing teams to expose this) to adopt a persona based development of the product, catering to the roles of cloud administrators, tenant administrators and end users.
After a short attempt to liase with Hadoop and a partnership with MapR - Nimbula went to support Openstack, opening their Director product to Openstack APIs.
Oracle in comparison had no similar offering, anything for orchestration across the private and public cloud. The work ahead will be to let Nimbula Director work and support Oracle's virtualization products. Moreover, once the ACME Packets acquisition will have happened - the IP will improve firewall throughput siginifcantly, which is critical for the performance of private / public cloud offerings.
Complimentary and / or Competitive Acquisition?
As we have seen Oracle may really need some cloud functionality from Nimbula. At this point in the speed of the cloud market, it's probably cheaper for Oracle to buy then to build that functionality. And the statement stresses the complimentary nature of the acquisition and indeed Oracle does not offer anything in the area of private cloud orchestration like the Nimbula Director product does.On the other side it does not hurt that Nimbula was a key RedHat partner (though RedHat put their chips down already, with the acquisition of ManageIQ, which left Nimbula in a somewhat weaker position. Likewise VMware was supported by nimbula - and VMware is closer as a competitor to Oracle given the focus of Oracle in the virtualization area.
OpenStack implications
Oracle was the last to join - via proxy of acquisition - the OpenStack party. Just two weeks ago IBM announced the support of OpenStack. Personally I don't think Oracle could stand out and wait longer. Nimbula now gives Oracle a board seat (with IBM, HP, Dell and VMWare - that will be a lot of fun meetings). This could be a bright future -- or the end of OpenStack. Remember UNIX - similar setup.Strategic value for Oracle
Oracle gets a very good private cloud product, with good orchestration capability across the public cloud. Through the OpenStack cloak, Oracle can even get visibility into machines and private clouds managed and run by their competitors. I am sure someone in the towers around Oracle Parkway is already building the Oracle Enterprise Manager extension to calculate the benefits of moving off other vendors hardware and one of Oracle's Exaxxx products. The low setup cost of Nimbula Director makes even a trial installation realistic.Advice for Buyers
This is good news for Oracle shops. Wait though till Oracle has a date for support of Oracle virtualization products by Nimbula Director. For non Oracle shops it comes back to trusting, that Oracle will keep the product open and running on OpenStack. Oracle would be foolish to say anything else at this point. But watch this area.MyPOV
A very strategic acquisition of Oracle, possibly at very low cost, given the challenges private cloud startups are facing. Oracle gains some key cloud talent, we will see if it will stay on board. And it's clear that Oracle really means to be in the cloud game and Oracle is in it to win it.I expect also less hyperbole around Oracle's False Cloud - as you would have to argue that an OpenStack compliant product isn't a cloud product. And that would make all OpenStack vendors False Cloud vendors. Ironically that's something that could happen, if the elephants at the table at the OpenStack board meetings play it out as it once happened to ... UNIX.