Year End Transitions

While I'm usually a "Happy Christmas to all and to all a good night" kinda guy I'm feeling a little Grinchy as the end of the year brings more news from the walking wounded of the storage vendor community. I've already written about MAID pioneer Copan's slide, but this week brings unpleasant news from Verari and the long beleaguered pair of Overland and Adaptec, with a ray of hope coming from Exanet who had looked pretty dead the last time we looked.

Verari's big idea was to mount their blades vertically so they could direct the airflow through a rack top to bottom rather than front to back like most servers. Using the rack as a chimney let them save power by using fewer fans. Of course that meant you bought Verari systems a rack at a time mixing and matching  Verari storage and compute nodes in the same rack. If you're building a hosting facility or HPC cluster with 500 nodes, buying a rack at a time is just about right. In fact, when Cisco announced UCS, I wondered if they were targeting large datacenters why they didn't build a blade chassis with more than eight slots. As server virtualization made blade server I/O more important Verari's systems just didn't keep up with UCS and HP's Virtual Connect. Despite the "we're not dead yet" protests by the few Verari execs still on the job, I can't see a recovery for them.

Better news comes from Israeli business papers that scale out NAS vendor Exanet will limp on another two months with their VCs, and co-founder Giora Yaron is kicking in another one million by court order. As I understand it, the board split over an OEM deal with Dell that, while it would have brought $20 million to the firm, would also have diluted current shareholders. Without that money, the firm couldn't pay its bills and a receiver was named.  The receiver says they'll run the company for two months with a for sale sign on the front lawn. They seem to be hoping Dell will snap up the company for its IP and developers for more than tickets to the next Spurs home game.

The big story is over at Adaptec where hedge fund manager, and major shareholder, Steel Partners, has pulled off a board room coup, replacing the CEO and putting the company up for sale. I can't say I blame them, as the past few years of Adaptec's history has been a unfocused mess that's bled off capitol in misguided ventures like iSCSI HBAs and multiple attempts to move up market from chips and controllers to storage systems, topped off by the purchase of Snap Appliance from Quantum and selling it to Overland a couple of years later at a substantial loss. Even worse were all of the attempts to expand into markets already occupied by their OEM customers, which also ended up costing them the bulk of the OEM business. In the '90s every server included Adaptec SCSI chips on the motherboard and many in the RAID controller. By the time U160 SCSI became the norm, LSI had replaced Adaptec with major OEMs like Dell and HP. Adaptec's sales have been sliding ever since. While I like Adaptec's latest generation of RAID controllers that use flash SSDs for read cache, I'm not sure I know who would cough up $400 million or more for Adaptec.  I've heard rumors about LSI and Xyratex, but I don't know.

Speaking of current Snap owner Overland, while this month's 1:3 reverse stock split raised Overland's stock price over the NASDAQ minimum of one dollar, the current price of around two dollars a share puts their market cap almost two million dollars below Nasdaq's $15 million minimum, so they're being threatened with delisting once again.  As a tape automation vendor, that despite many attempts hasn't gotten any real traction with their disk based products, I don't see Overland making a big comeback. In fact, they've reached the stage where my concerns about their ability to stay in business are leading me to recommend alternatives even to clients that have been happy Overland customers in the past.