The OpenStack initiative has been the topic of extensive discussion since its inception. OpenStack has seen good adoption as a platform for science and education, and by some webscale organizations wanting to build their own private clouds. But as a commercial proposition, it has been questionably successful.
The large number of venture-backed startups that were set up to deliver commercial OpenStack distributions have largely been acquired or shuttered, and the remaining player, Mirantis, has moved on to broader offerings. HPE, one of the strongest historical backers for OpenStack has given up, selling its OpenStack business off to SUSE.
And so I was intrigued, and a little dismayed, when a source pinged me recently to give me the low down on some rumored changes within IBM. Regular followers will know that IBM has been a long-term supporter of OpenStack, and a few years ago acquired one of the more well-known OpenStack service providers, BlueBox. With HPE’s move away from OpenStack, and a hugely de-emphasized involvement by other large vendors such as Cisco and Intel, IBM was one of the only remaining supporters.
My source informed me that IBM has moved most of the old Bluebox team onto other projects, only leaving enough staff to service existing customers. They also told me that internal IBM teams are being told to move off of their own Bluebox clusters. This is particularly interesting given that Jesse Proudman, the founder of Bluebox, and a key figure in the OpenStack community, was seen as a key OpenStack evangelist within IBM. My source tells me that Proudman has been doing other work inside of IBM for quite some time now.
There is little visibility as to whether these moves indicate an intention to sunset OpenStack, or whether it is simply a chance to ramp up the revenue it generates, although my source suggested that internal factions inside of IBM that didn’t want Bluebox/OpenStack to be successful won at scuttling it. These factions apparently wanted VMware, and recent moves suggest that the virtualization giant won out in the end.
I’m hearing a conflicting rumor that suggests IBM has been meeting with Canonical representatives in order to investigate a partnership whereby Canonical OpenStack is offered to IBM customers. This arrangement would mirror the HPE/SUSE deal whereby HPE gets to offer an OpenStack product on their product sheet, but without having to actually do any engineering work themselves. It would also parallel an existing relationship between Red Hat and IBM. A way to, sigh, sell more boxes, it would seem.
I have reached out to both IBM and OpenStack for comment on these rumors, and an indication of IBM’s intentions going forward. For its part, an IBM spokesperson replied saying:
we aren’t going to comment on rumors at this time
So, not much from them. Jonathan Bryce, the executive director of the OpenStack Foundation, was more forthcoming, telling me that:
We’ve seen consolidation of OpenStack distributions with HPE partnering with SUSE and IBM continuing their long term partnership with Red Hat. This is a good thing to happen as the market matures and the choices become more clear for users and businesses.
MyPOV
HPE and IBM. Birds of a feather. They lack the ability to innovate and seem to be happy to ignore the very real disruption which is coming down the line to them. At least, I guess, HPE has the decency to asset-strip itself and sell off all the good parts before the end comes. IBM just continues to ride perennial quarters of lowering fiscal performance.
For the OpenStack Foundation, it is, of course, some unwanted bad press. To be honest these legacy vendors were never going to be the future for the Foundation but the optics aren’t ideal. As we’ve seen historically from the companies that figured it out with Linux, these same companies are applying what they learned to new open source communities more successfully than the companies with a hardware tradition.
