NetApp's glass-house-stone-throwing exercise
Innovation: all is not butter that comes from a cow
Comment Innovation, eh. Don't we all love it? But what is it?
This thought was spurred by reading a blog by NetApp's Matt Watts, UK-based director for technology and strategy, in which he criticized Pure Storage for not being innovative.
Pure started with a good product ... but from that initial product how much incremental innovation has there been?
Replication? There's no way you can claim this as innovation. This is a basic and fundamental feature that any enterprise array simply has to have.
Evergreen? A good idea and a well packaged proposition but that's it, no ground breaking technology or innovation enabled this.
New hardware? All storage platforms get new hardware over time. Sure there are some interesting physical characteristics, but hardware is there to deliver performance and in that perspective it's not set the world alight.
There doesn't seem to be much else.
Here is his list of recent NetApp innovations:
- Huge performance and efficiency optimisations in ONTAP for All Flash FAS
- Cloud ONTAP for AWS
- AltaVault for Cloud Backup
- StorageGRID for Object Storage
It seems to me that to claim innovation though having bought a cast-off Riverbed product (AltaVault) or a second/third tier object storage supplier (Bycast and hence StorageGRID) is a tad rich.
To optimise a legacy disk array operating system to better use flash is large potatoes in NetApp's universe, ONTAP being crap before, but small potatoes elsewhere. Every other mainstream array supplier has done the same thing.
Is CloudONTAP innovative? Its NetApp's array software running as a VM in Amazon's Elastic Compute Cloud. That seems pretty average. Getting your storage array software running on its own in a server and then running it in the public cloud seems pretty straightforward to me.
Look at this from Pure's point of view. In its recent announcement, its two controllers have high availability via NVMe PCIe non-transparent bridging, and Pure is implementing NVMe connectivity directly to the flash modules, thus increasing performance. Pure also gets itself a route to NVMe fabric linking of its arrays to servers, thus banishing network latency.
Has NetApp done this? Have other all-flash array suppliers?
An innovation is a new idea, thing or process that has an impact beyond the immediate area in which it happened. It's not just novelty or incremental product or business process tweaking that enables one business to catch up with another.
Innovation has been asleep at NetApp. It has embraced somnambulism as a product development strategy for some considerable time.
If it hadn't then the all-flash array, hybrid array, software-only storage, and virtual SAN/hyper-converged infrastructure appliance startups wouldn't have had such fertile ground on which to grow.
NetApp has had many innovation failures and sagas. Shall we mention FlashRay? Spinnaker taking many many years to be integrated? The resistance of NetApp's customer base to adopt the resulting Cluster-mode ONTAP? The Topio acquisition?
It's all very well to gloat at Pure's discomfiture over the Gartner revenue numbers but people in glass houses shouldn't throw stones.
Or, as the Jewish saying has it, all is not butter that comes from a cow. ®