Back in 2006, when Sun announced project Blackbox, they were doing something revolutionary. Taking the ubiquitous standard shipping container, Sun was outfitting it as a full-blown datacenter, with the capacity to offer strategic computing services anywhere a container and power could be delivered.
Fast forward a half-decade and datacenter containers are an expected part of the product lines of the world's datacenter hardware vendors. Not high-volume items, but often showplaces for the latest technology these vendors want to provide to their datacenter customer.
But for Sun, now just a part of the Oracle software empire, the Sun Modular Datacenter, formerly known as Blackbox, is just a memory, albeit one that a customer can purchase, if that memory is strong for them. Especially since the Blackbox will no longer be a cataloged item. No containers will be stocked, but if a customer wants one, Oracle will build to order.
This begs the question of whether there is a serious market for containerized datacenters, or is it just a reflection of the fact that Oracle, even after the acquisition of Sun, remains a software-focused company without the understanding of the datacenter hardware market necessary for success at this time.
Oracle could have gone two directions with the modular datacenter model; embrace it, offering the on-demand Oracle in a box, with a drop in datacenter preconfigured to the customer's needs to run the Oracle database, as a module in an existing datacenter, or do what they seem to have done, brush the hardware side of the business under the rug and hope that they won't need to deal with developing a competitive modular datacenter offering.
Regardless of the motivation behind the Oracle handling of Blackbox, I don't see how this can be positioned as a positive choice for the Oracle customer.