Most people have been expecting this for a long time, and have wondered why it didn’t happen sooner, with Oracle ditching HDS as an OEM partner almost immediately after acquiring Sun.
I have read, and heard over the past year that Oracle has been for the most part destroyed in the storage market (servers doing badly as well) as a result since their Sun storage products just are not competitive. Many larger customers have been leaving to the likes of HP and IBM who could offer the “one stop shop” for servers and storage (even before HP bought 3PAR, HP had and still has their OEM’d HDS equipment).
In some informal talks with some HDS folks last year they seemed quite happy that Oracle was no longer an OEM, saying that the people over at Sun/Oracle weren’t competent enough to handle the HDS stuff (*cough* too complicated *cough*), and so HDS just went in direct with most of those customers that Oracle walked away from.
Finally someone at Oracle woke up and realized there still is, and will continue to be for some time a big market for traditional SAN systems, far bigger than the market of customers willing to risk putting their data on cheap SATA controllers on servers running ZFS with high failure rates and poor performance.
So it finally happened, Oracle is buying Pillar. At first look however it really does seem like an odd scenario, from their SEC filing –
The Earn-Out therefore will only be paid to Mr. Ellison, his affiliates and, if applicable, to the other Pillar Data stockholders and option holders if the Net Revenues during Year 3 of the Earn-Out Period exceed the Net Losses, if any, during the entire Earn-Out Period.
There’s no specific mention whether or not Larry is going to pay himself back for the $500M+ in loans he has given to Pillar over the years, so I suppose not. In any case it won’t be until the end of 2014 when we might discover what value Oracle has placed on Pillar. One commenter on The Register mentions Pillar’s revenue as $29M per year, don’t know where that came from though, doing some searching myself I found references to roughly $70M in revenue, to $3B in revenue (if that was the case they would of IPO’d)
I think it’s a good deal for Pillar to, they get much better validation on their products in front of customers.
I’ve gone through quite a bit of the information on the Pillar web site and to-date I have not seen anything that would make me want to buy their product, and have yet to hear any positive words coming from the people I know in the street/industry (granted my community is limited).
But it sure as hell beats anything that Oracle has been offering their customers recently, that alone may be enough to drive a decent amount of sales.
Pillar posted some updated SPC-1 numbers recently, a significant improvement over their original numbers, though nothing ground breaking from a competitive standpoint.
In other news, two early social media giants have fallen – MySpace being acquired for $35M, and Friendster re-inventing itself as a gaming site with Facebook authentication. I’d bet the infrastructure behind Myspace is worth about $35M by itself – Newscorp really wanted out!
Oracle is primarily a software company and its incompetence in hardware business is why the acquisition of Sun has been a disaster. It had a golden chance to fix Sun’s hardware business and is just killing it with incompetence. (Numerous posts refer to 3-6 month delays in getting support contracts renewed). Given that, the Pillar acquistion will also most likely fail. Do you see existing Oracle/SUN customers moving to Pillar from Storagetek or HDS? I don’t.
ZFS still has a lot of potential, IMO.
Suresh
Comment by Suresh Rajagopalan — July 7, 2011 @ 10:48 am
Thanks for the post!
Yes I agree most people in their right minds would not use Pillar over some of the competition, but I do believe Pillar sales will go up quite a bit with Oracle pushing them on their customers (because there’s so many people not in their right minds out there), discounting and “optimizing/integrating” enough to convince some people to buy Pillar.
As for ZFS I agree it’s good technology, it’s the pairing of ZFS with shit hardware is where I have the issue. Put ZFS on good HA hardware to get most of the best of both worlds, fancy software features from ZFS and high availability from the underlying storage. Of course the ZFS data integrity functionality is out the window once you abstract the storage with hardware RAID on the storage array, but that’s kind of unavoidable. If you haven’t seen the post that is related to this from last year see – http://www.techopsguys.com/2010/09/09/availability-vs-reliability-with-zfs/
Comment by Nate — July 8, 2011 @ 8:37 am