Storage start-ups fail to set the world on fire
How IT fell for file storage growth myths
Not a general problem
What these four groups of startups didn't realise was that the generality of businesses would be content - not happy maybe, but content - to carry on as before. NAS boxes grew to hold more data and it was easier to carry on with existing file storage processes than to bring in new and untried ideas like, Copan's MAID, Isilon's clusters or Acopia's file virtualisation or whatever.
They also over-stated the influence of compliance and e-Discovery. The sell here was basically one based on scaremongering: look how much customer X got fined because they couldn't find a file in time! Not enough people were frightened into buying these start-ups' products, and the existing filer vendors did enough to assuage the compliance and e-Discovery concerns of their customer base.
The quartet of file storage problem solving startups then met other problems. Sub-file level deduplication removed a prime reason for having their products, especially when they worked within the existing backup software and process infrastructure: witness the success of Data Domain. Specialised dedupe vendors built and developed products, and grew their companies faster than the group of four file storage problem solvers we are looking at.
In the last couple of years, EMC and IDC have done a fine job of alerting everyone to the storage consequences for the digitisation of media, of social interactions and increasingly of intelligent device communications. The file storage problem pioneers are being proved right, but it is not doing them much good.
They came into existence on the back of an exaggerated problem. Because of this, they weren't generally able to capitalise on the limited demand for their products and build sustainable businesses before the major storage vendors did enough to catch-up and take their market prospects away. That old consultant's mantra of get big, get niche, or get out (or bought) applies here.
None of them have got big. File virtualisation is a dud. Persistent storage is a nice feature to have with an archive, but of limited appeal on its own. Clustered scale-out filing is a niche, with major suppliers ready to capitalise on it if it becomes mainstream. Archive is a niche with cloud storage positioned as a probably valid alternative storage choice to whatever drive arrays you choose.
This piece started with the statement that there is no general large scale file storage problem. I'd contend that there still isn't. The problem area is bigger than before but there's no sign yet that business in general needs a massively scalable, logically single file store, with a global namespace, compliance and e-Discovery facilities based on the greenest possible drive arrays.
Some businesses need some of this. Not enough businesses need all of it, and that's why not one of the file storage problem start-up companies mentioned here have become successful and major suppliers in their own right. ®
Sponsored: RAID: End of an era?