NetApp is one of those over-night successes that has been a long time in the making. It's been steadily building on its networked-attached storage (NAS) filer base for several years and is now number two behind EMC as an independent storage system vendor.
It is now about equal to EMC on NAS shipments and has a commanding leading in iSCSI product shipments through having adding iSCSI capabilities to its products. What NetApp didn't have until recently were enterprise-class datacentre credentials, arrays and management software. Nor does it have the ability to combine heterogeneous NAS products into a single highly-performant global namespace.
Four weaknesses some would say. Three of them have been all but fixed. The answer to the fourth is in a prolonged development phase.
Through a reselling stroke strategic alliance with IBM it has IBM's salesforce carrying a NetApp product message into the world's top organisations, the ones with IBM gear in their data centres.
Both EMC and NetApp are in each other's core storage market. EMC has its Celerra NAS and NetApp now has its FAS 6000 enterprise drive arrays. These can have up to 1,000 hard drives, totalling half a petabyte or so of capacity, and take NetApp up into the Symmetrix array area. It's surely no coincidence that a scant time before NetApp's FAS6000 announcement EMC upgraded Symmetrix to have slightly over 2,000 drives supported. EMC Red Indian, he say: "Me stay top dog, you nowhere," so to speak.
NetApp has very recently signed an agreement with Symantec to supply a co-branded version of Symantec' CommandCentral Storage heterogeneous datacentre SAN Management software.
NetApp is generally perceived to not have a multi-vendor NAS virtualisation capability nor an ability to cluster distributed NAS systems together. Because of this its NAS growth in terms of file space and general scalability is perceived to be limited. Yet it does have NAS virtualisation capability with the V Filer and Data ONTAP 7G. It's NAS clustering capability is in beta test with Data ONTAP GX.
The incredibly fast-growing MySpace website business recently and publicly chose Isilon for its extreme high-end NAS needs over NetApp citing growth and performance reasons.
The reason this happened illustrates both NetApp's strength and a weakness of its approach. Actually 'weakness' isn't the right term here. It certainly delays the time before new product technology comes to market and it's a weakness in terms of NetApp's customers being receptive to pitches from competing vendors but a stronger product is the result. It certainly has been so far.
All NetApp products share the same operating system software: Data ONTAP. This means that when NetApp adds a capability, such as having its drive arrays serve blocks to Fibre Channel or to iSCSI, then Data ONTAP has to be upgraded and regression tests carried out to ensure that existing functionality isn't compromised.
As each new capability is added to Data ONTAP the software development and testing burden becomes greater. This is not the case for, say, EMC, which has acquired numbers of products and technologies that don't share a common architecture and operating software environment.
It means that EMC can bring products to market faster than NetApp but they are not so well integrated into the rest of its product range. NetApp products are well-integrated together and customers get the benefits of that with better inter-working between NetApp products.
EMC can buy a technology company and introduce products quickly. NetApp can buy a technology company and then has to assimilate its technology and produce a new version of Data ONTAP embodying it. This is slower than EMC's approach. Thus NetApp's ability to capitalise on its Spinnaker NAS clustering technology buy has been delayed because of its approach. The resulting product should be a very good one but it won't be first on the market.
Also, with the GX product NetApp now has two versions of Data ONTAP which will, eventually become one product. Clearly, software development is complicated for NetApp.
It means Montilio, Isilon, PolyServe and Acopia can all make strong sales pitches to NetApp customers needing more NAS performance and scalability. However, NetApp has closed off a susceptibility for its base to buy high-end arrays from other vendors, such as EMC Symmetrix, by introducing new high-end product.
The long haul
You don't build enterprise-level datacentre credentials over night. It takes years. NetApp has started out on that road and is hoping that, as with filers, it will progress to the top table, along with EMC, HDS and IBM. Surely IBM and NetApp will develop their agreement so that a single or combined product set emerges.
We will have to wait and see. What NetApp isn't doing is branching out from its storage roots and going into virtual application server environments (VMware), document content management (Documentum), data backup (Legato and Dantz) and larger-scale professional services as EMC is doing. Perhaps EMC's growth will slow as management focus gets spread thinner and thinner. Perhaps not. NetApp's management focus is all on storage. Its track record of introducing and developing storage technology is excellent. Let's hope that, for its sake as well as that of its customers, its data centre strategy and products provide it with Growth ONTAP.