Pillar Data - Larry Ellison's other storage company
Life after Sun
Analysis Oracle buying Sun Microsystems has caused people to wonder about the future of Pillar Data, Larry's other storage company.
Pillar Data is backed by Larry Ellison's personal investment vehicle, Tako Ventures, possibly to the tune of half a billion dollars.
Founded in 2001 the Silicon Valley-based firm has developed the Axiom storage line, guaranteeing 80 per cent utilisation, which is, it says, twice the industry average. Better utilisation means fewer storage systems. Axiom is also "application aware", which makes for much more efficient storage provisioning, according to Pillar.
Oracle has its own hardware ventures, notably the database machine developed in partnership with HP. And now it is buying Sun, which has its own significant storage offering.
So where does Pillar fit into Larry Ellison's scheme of things? Come the end of the recession, he can make a decent return on his investment by getting the company to IPO. Or, he could broker a deal one day to fold Pillar into Oracle.
This is the more intriguing possibility. And strategically, it makes sense. A view from inside Pillar is that the Oracle boss is personally funding Pillar to see if, among other things, if it is possible to build a storage offering that sidesteps third party storage vendors.
If Oracle owned the entire stack, the whole works from disk platters through to its own software then customers avoid the sometimes great expense of, say, an EMC or NetApp system capable of operating and storing the data needed.
So is Sun the answer? The answer in the Pillar camp is "no".
The Sun arrays
Sun's storage arrays include high-end USP-V HDS products that Sun resells to a different customer base from the mid-sized enterprises that tend to buy Pillar Axiom arrays.
Moving on to Sun's mid-range arrays, which contain technology inherited from StorageTek, the company bought by Sun for $4.1bn in 2005. These are based on products supplied by LSI or Dot Hill and are, loosely speaking, first generation SAN arrays, and not second generation iterations, as exemplified by 3PAR, Compellent and Pillar.
These StorageTek products are storage array platforms without the ability to offer application-specific quality of service or utilisation guarantees. Combine these limitations with the need to pay LSI and/or Dot Hill for the supplied enclosures and controllers, and the Axiom arrays look less costly and better suited to work in many Oracle customers' environments. So says the Pillar camp.
Now consider the Sun Open Stage 7000 products and the X4500 Thumper hybrid server/storage arrays, which are produced by Sun's system division. A Pillar aside is that it seems that whenever server guys build storage they always build servers.
The Sun 7000 business model, with its commodity hardware aspects, is good news in Oracle-land, but its open source software is not. Not to Oracle, any way. Neither is the bag-of-bits aspect of the 7000's software environment particularly attractive to many customers, according to the Pillar camp. Here are the software lego blocks; now build the storage array hardware and software system yourself. No thank you. I want to have an easier time implementing my storage array. (N.B. Note the reader comments that attack this characterisation of Sun's 7000 line.)
Combine that with the lack of any application-aware quality of service facilities and utilisation guarantees and the Pillar Axiom stacks up well against the Sun drive arrays. If Oracle bought Pillar then it's thought that the StorageTek drive arrays would quietly wither away, The high-end HDS ones kept, and the Sun X4500 and 7000 products also undergo withering with the Axiom surviving and prospering, perhaps with added Sun technology, such as ZFS.
The upshot is that Pillar is not threatened by the Oracle-Sun deal. But its future might be made clearer by it, with the Sun acquisition perhaps functioning as a catalyst to make developments happen regarding Pillar.
Nobody knows what's in Larry's mind, except Larry and his close colleagues, so this analysis is not informed by a close understanding of his intentions. Hopefully, it adds some insight into what could happen with Pillar in coming years. ®
Sponsored: DevOps and continuous delivery