Original URL: http://www.theregister.co.uk/2011/07/04/oracle_pillar_requiem_and_redemption/

When Oracle bought Pillar for £zilch

Is Pillar's fate its opportunity?

By Chris Mellor

Posted in Storage, 4th July 2011 12:18 GMT

Comment It was a telling omission when Oracle introduced a storage guy to talk about why Pillar Data was such a great acquisition. It wasn't Mike Workman, Pillar's CEO and founder, but Phil Bullinger, Oracle's SVP for Disk Storage.

Contrast that with Dell parading Phil Soran, Compellent's CEO, when it bought that company, or HP showing off 3PAR's David Scott when it bought his company. Mike Workman was banished from Oracle's stage.

The history of Pillar

Pillar is Mike Workman's baby. It's a carefully crafted product with Workman's hands-on management style all over it, including in its marketing and sales. He's a no-bullshit, hands-on chief exec with the down-to-earth style of an experienced disk storage engineering guy who's been around the block a good few times, some with Big Blue, and knows the score.

Typical of this was his insight that, yes, you needed to short-stroke drives for performance, putting high-access data on the outer tracks of a drive where they got the fastest access, but there was no need to waste the inner tracks: you could put low-access rate data there, getting a nice balance between speed and capacity. No waste that way and elegant engineering.

The whole point of storage is to deliver data, stored cost-effectively, to applications in a dependable and appropriate way. So he developed Axiom array software to do just that, with storage quality-of service levels and application-aware storage profiles specifying how fast apps got data and how it was protected. Customers liked that focus, that sense that Axiom was a data storage tool designed and built by a man who knew about storage tools.

Workman is a polymath. He appears to have few problems with the work/life balance issue, being a family man, a skilled underwater photographer, a respected amateur fireworks technician, and the operator of his own winery. Oh, and by the way he runs, he ran, a storage company.

Mike Workman wine label

Label from Mike Workman winery.

Stencilled on the flat roof of Pillar's San Jose HQ is a silhouette of Clams, his weiner dog (miniature dachshund). Workman is far, far removed from the image of the young entrepreneur chasing the fabled Silicon Valley start-up dream, and some would say the dream has eluded him and his company.

Good timing

The timing was good for Pillar's creation 10 years ago. There were other storage startups adding their own secret sauce, both software and hardware, to make disk arrays do more yet cost less and be easier to manage. Workman's bad luck was to end up competing with two companies that did it better in their markets. Both were run by experienced guys too and had, it has been suggested, better all-round executive teams than Pillar.

David Scott was the British CEO of 3PAR and it had a hardware-accelerated InServe array positioned as a very, and I do mean very, easy to manage high-end product that was popular with IT service providers. Scott focused 3PAR on this and had effective marketing plus he focussed a lot on the revenue bought in by his regional sales people and their reps. Quite a few were shown the door. Scott was a skilled champion for his company with a mischievious and scalpel-sharp way of talking to customers' execs about the competition, namely EMC or, as he put it, Emit More Carbon. This was not something Workman, with his more down-to-earth approach, could emulate.

There was no effective way Pillar's Axiom could match 3PAR's high-end box. Nor could it match the box produced by Phil Soran, Larry Aszmann and the other members of the founding team that built Compellent's Storage Center into a devastatingly effective array for the mid-range. Compellent competed with CLARiiON and 3PAR with the DMX and both became the leading SAN competitors to EMC in these areas, leaving Pillar to pick up crumbs in the sparsely-populated ground between these two great companies.

And crumbs was all Pillar got. While 3PAR and Compellent powered their way to multi-million dollar acquisition exits for their VC backers and founders Pillar struggled to get traction. In 10 years it sold just 1,500 systems to around 600 customers. These customers loved the product and the company; there just weren't enough of them.

Oracle says it has 20,000 customers for its Engenio-based SAN array, now sourced from NetApp, and will continue to sell and support [the Engenio arrays] for the foreseeable future. There are more than thirteen Oracle Engenio customers for every Pillar customer.

The truth of it appears to be that Mike Workman and his team did a fine job, but 3PAR and Compellent did an even better one.

Pillar, acquired for nothing, not one dollar, not a cent.

Worth nothing, now

So that's Axiom, a general purpose SAN and NAS unified array with application-aware quality of service, and there's Pillar, acquired for nothing, not one dollar, not a cent. After 10 years hard and dedicated graft, the net result is apparently worthless. Pillar is loss-making and Larry Ellison has gambled away half a billion dollars on what some have characterised as a vanity storage toy.

That's accounting for you, and the need for Oracle to appear financially whiter than white, and not be seen as Larry's company bailing out Larry in a private venture that has gone sour. There was no other buyer on the scene and this was the best, in fact the only, offer on the table. Larry's toy could have been tossed out of the pram altogether. As it is, it lives on to fight another day.

Pillar, it can be argued, is a work in progress and actually has a lot of potential. If it turns itself around from its current loss-making state and makes enough revenue in three years time to exceed its losses between now and then, its stock-holders will get their share of three times the difference. There's a chance there for serious money.

The only route to it though, is by doing what Oracle wants: building a block storage array that runs Oracle software better than the competition. The Axiom's future is to be the storage layer in an Oracle integrated, block-access stack, with every major part of the stack being supplied by Oracle for Oracle, not for SAP, DB2, SQL Server or any other software application supplier competing with Oracle. It will store other applications' data but be tuned for Oracle software.

Axiom versus CLARiiON CX4

Pillar Axiom vs EMC CLARiiON CX4.

Oracle is positioning Axiom against EMC's CLARiiON and says it's better than the CX4. It is getting columnar compression and will have its application profile capability extended significantly.

Axiom has to become the best SAN storage for Oracle software. That's its management's sole focus from now on. Will its current management be left to get on with the job? Does it want to, after what some could perceive as an insult delivered to them by the ruthless and possibly careless Croesus running Oracle?

Will Mike Workman have to report to Phil Bullinger? There are bound to be efficiency savings as Pillar's back office functions get absorbed into Oracle's, and as other responsibilities inside Pillar are combined with or integrated into the equivalent Oracle ones. This could affect finance, sales, channel sales, marketing and HR. Not insubstantial cost could be saved that way.

Do Mike Workman and his team have the energy and enthusiasm left for another three years hard graft and grind, doing it Oracle's way and not their own, but with the prospect, by November 2014, of actually realising the Silicon Valley dream and making it big time? Can they see themselves pulling this off, vindicating all the work they have done so far, getting their serious walk-away money, ensuring their reputations are flavoured with success and not tainted with failure?

Can they do it? Can Pillar redeem itself? Let's hope so. It would be a sweet, sweet reward, and tomorrow's greenbacks would be compensation for today's brickbats. ®