Copan plugs on with its creaky green revolution
What happens if you solve a problem no one cares about?
Copan makes the greenest disk storage going, with no exception. So with everyone running out of power and data centre space, drowning in unstructured file data - why aren't customers buying it in droves?
The trouble is that Copan's products solve a problem many customers don't know that have and aren't motivated to solve anyway. Also, EMC, HDS, Nexsan and others have queered Copan's pitch by marketing their spin-down products as power-saving fixes for storing little-accessed data, even though they only save a fraction of what Copan can.
Copan's core technology is MAID, or Massive Array of Idle Drives. Three-quarters of the maximum 896 drives in a Copan Revolution array are switched off - they don't consume power nor generate heat. That means Copan can do something no other supplier does, which is to pack them so close together that if they were all switched on at once the array would overheat.
When data on idle drives is needed, then the drives are spun up so users get their data in seconds and don't have to wait for a tape reel to be located, mounted and streamed. The trade-off is slower but still online access against vastly lower energy bills and radically reduced floorspace needs.
Chris Santilli, Copan's chief technology officer (CTO) and a co-founder, says that a customer needing an array to store 250TB, an average Copan deal size, "would need five Nexsan cabinets, but it's only half a cabinet for us".
With 1TB SATA drives Copan can pack 896TB into one and a half data centre floor tiles - ten square feet. Not enough people care, though, in the great scheme of things, because most CIOs are not responsible for energy budgets. They might weep for the planet, but these are crocodile tears as far as data centre IT purchases are concerned.
A Copan survey of 250 IT professionals in Fortune 1000 companies found that 70 percent said energy efficiency was a priority for their company. That was well-meaning green fluff. Of these greenies, 66 percent say energy efficiency is considered but not the top priority.
Another 15 per cent said, confusingly, that energy efficiency is not important, leaving less than a fifth, 18 per cent, thinking it very important in storage buying decisions. That's around 13 percent of Fortune 1000 businesses, if we do a straight extrapolation, who would be ready to buy a Copan box if they knew they had a problem with old and little-accessed data clogging up their primary arrays.
But most CIOs also don't actually know that their primary arrays, meant to store transaction data, are actually chockful of data nobody wants to look at much. Santilli calls this persistent data, the stuff you have to keep and can't afford to send to tape vaults because you might need to look at it some time and it has to be online.
There's no way of measuring the stuff unless you monitor file access patterns. ESG analyst Mark Peters, roped in by Copan to talk about the survey, said: "Although no one truly knows the actual percentage of data that is persistent, we estimate it to be at least 60 per cent of data, probably more, and definitely growing." Gee, thanks Mark, you don't know either.
To reiterate - Copan sells a product solving a problem most CIOs don't know they have and aren't incented to solve anyway.
What's Copan doing to be more successful? It's building alliances with HSM (Hierarchical Storage Management) suppliers such as Sun (SAM XFS), Symantec (Storage Foundation VXFS) and Quantum (StorNext) whereby users see a single logical file space encompassing transaction data on primary arrays, persistent data on Copan's Revolution arrays, and vaulted data on tape.
That makes the Copan kit easier to fit into a data centre environment, but it's still Copan and its channel partners mostly doing the selling, and selling to CIOs unaware of the problem and heedless of energy budgets.
Copan is improving its product. The arrays can be presented as virtual tape libraries as well as a straight file store, and they also support block-level deduplication with a FalconStor partnership.
The future will see the arrays getting a substantial boost in processing power - we're in a quad-Xeon era after all - and the arrays will then be able to run deduplication, encryption, replication and other functions themselves.
Santilli is keen on XAM as a way for open source developers and others to produce content-addressed storage (CAS) software which could run on the Copan array controllers. He is a great believer in leverage. However, that hasn't extended to getting Copan products supplied by EMC, HDS, HP, IBM or NetApp. Of the major IT system vendors only Sun Professional Services has a reseller agreement.
Revolution in waiting
Copan is privately-held and profitable, and has been through four funding rounds bringing in $89m. It has at least 250 installations in 17 countries, probably more, and is trading profitably. But it isn't a superstar, and it could be.
It seems likely that any Copan sales growth spurt depends upon CIOs getting energy budget responsibility, and on realising that their transaction arrays are storing data that has no business being on them, compliance data or what used to be called reference data. There we have it; the world's greenest and densest storage array, ready for a market with customers running out of energy and choking under an onslaught of file data.
It's not making as much headway as it could because customers aren't organised to recognise the problem and sort it. They are not hurting enough.
Santilli says: "A start-up has to solve a problem. If you're not solving a problem you have no right to exist." Copan is solving a problem, a big one, but it could be so much more successful if customers were feeling their energy cost and file growth pains more. Its product is revolutionary, it's just that customers aren't ready for the revolution - yet. ®