NetApp's unified storage platform is a compelling vision for a customer; one platform to support pretty much all your storage needs. It is a powerful sell, it is still pretty much a USP for them; everyone else has to fake it by glomming together storage products and pretending.
But if we dig a little deeper; is NetApp's unified storage platform going to become a millstone? I suspect that it might; arguably it already has. Just looking at the length of time it took them to get OnTap 8 out of the doors and the issues that has brought; the bringing together of GX and traditional OnTap took too long and probably depleted them of development resources.
It might have been better if they had decided to let them live as two separate products rather than the painful union that faced them. Concentrating on making OnTap 8 64 bit and ensuring things such as seamless migration from 32 bit aggregates to 64 bit aggregates might have actually been of more value to more customers.
Having to integrate any new idea/innovation into OnTap slows time to market because time needs to be taken to work out how to technically integrate it; how to test it and generally make sure that it does not break existing functionality or at least have detrimental effect.
The competitors can 'simply' build a new product line without impacting their existing code-base; yes, they can borrow from the existing base and utilise common components; for example the underlying base operating system may be the same custom Linux environment across their products but they do not have to worry about detrimental impact.
And now we have object storage; is integrating object storage just a step too far for the USP. Actually, technically, it should not be a huge challenge but commercially I can see issues. NetApp's object storage is going to be the most expensive object storage in the world! They are going to be competing with commodity disk prices and the only way that they can do this is to trash their own margins.
I think that is going to be painful. it might well be better for NetApp commercially to be able to sell OnTap ObjectStore (not it's real name, I just made that up) as a completely different product but that would mean accepting that the Unified Storage Platform is not the best answer to what may well be a purely commercial problem. Technically and aesthetically it is very elegant but it is it long term practical?
Of course, EMC still have far too many storage products and their sales-team live in a general state of confusion.
The dilemma I face with this, is although it maybe a Unified storage platform, it doesn’t offered Unified storage management, yet alone Unified information management.
It all relies on the premise that the NAS is the storage. But lets not delude ourselves. NAS is one piece of the equation. Data resides in other locations and NetApp’s reach on application servers, workstation and laptops is weak.
We may scoff at EMC disjointed list of what almost appears random products, but they have at least acknowledge that storage and information management doesn’t end in the big iron.
I think NetApp need to look at the bigger picture.
Unified my eye. Unified as long as you stick with NetApp’s (or any other storage vendor’s) hardware platform. Unified, as long as you commit to NetApp’s proprietary O/S and file system – meaning no ability to implement best of class storage apps from 3rd parties. Unified, as long as the solution to all your storage problems is NAS.
There are new solutions available that use open source file systems and are hardware and network agnostic. A great example is Nexenta Systems’ solution. (shameless plug) NexentaStor runs on any industry standard architecture server and connects to anyone’s storage and utilizes open source ZFS file system.
The days of proprietary storage hardware running proprietary software are coming to an end.
Interesting post Martin. One size rarely if ever fits all. Eventually as you enter more markets it becomes jack of all trades and master of none. The monolithic nature slows you down.
Even NetApp realise it. They were about to buy DataDomain and that would have been a different platform.
Working at HP I won’t pretend we have one platform for all of your storage needs. I will even say we have too many for any single customer. However with the portfolio we have including our D2D platform & the recent acquisitions of LeftHand NetWorks and IBRIX, I will contend that we have all we need to build a converged infrastructure that can scale compute, capacity and connectivity independently to any workload. No Compromises!
Lee Johns
HP StroageWorks
Hi Martin,
Have a look here as I collect my thoughts on this topic: http://blogs.netapp.com/exposed/2009/08/the-burden-of-success.html
@Fab – we will be disclosing much more detail on our Cloud business and technical strategy over the next few weeks. In the meantime, based on your comment, I encourage you to explore our Provisioning Manager solution for a glimpse into Unified Storage Management at scale.
@Tom – competition is great, and the FOSS movement is performing the necessary Darwinistic role of keeping non-FOSS value-add vendors like NetApp on our toes!
@Lee – please have a look here: http://blogs.netapp.com/exposed/2009/06/geneouses—hetero-and-homo.html
Val, I still think that you have a problem in the bulk storage arena. You have competitors who are shipping product at a much lower cost than yourselves; perhaps this is a market that NetApp do not want to be in.
Although I have blogged on the consumer market for Cloud; it is not merely this market which you might miss out on. There is a market for storage which does not necessarily come with all the bells and whistles of a NetApp solution; this may be because the functionality is delivered inherently as part of the application.
For example if you look at Digital Asset Management and digital workflow; much of the traditional capability of a storage array is encapsulated in the application. The richness of a traditional array feature set is wasted and an unwarranted expenditure but it is still required to be shared storage.
And then there are products like ZFS which could also over time replace much of the functionality of an array.
Technology is an interesting space, a product can grow from nothing, succeed wildy and then vanish without a trace. Track-record counts for very little.
Yes I agree with Martin’s comment concerning track record. To add also, track record is all well and good if your track record is one that demonstrates the ability to adapt. I’m not seeing this from NetApp of late.
I’m not plugging EMC here, but they recognised that a long term big iron hardware play was limiting. They have since adapted to offer solutions across the spectrum of storage and information management and more importantly, started to cater for a big range of customer budgets.
NetApp ironically seems to be moving to where EMC has tried to de-emphasise from.
In a financially sensitive climate where IT funding is getting slashed, people are looking for affordable solutions with enterprise grade functionality. Requirements havn’t got smaller, on the contrary, but budgets have!! Kudos to the likes of Nexenta for delivering this by the way. They remind me of a young NetApp going after EMC business.
(Of course HP have for a good while been providing solutions for varying budgets and I’m not including them or IBM here as both have always had broad portfolios).
I suspect that NetApp is riding out this recession on the back of professional services rather than any uptake on hardware sales. But you can only keep telling your customer base how to make better use of their hardware for so long.
Is NetApp jumping on the cloud bandwagon going to be game changing? It’s likely to be a “me too” play. (Unless they do a Cloud staging hybrid , no big vendors have delivered this yet)
I think NetApp’s tight relationship with Symantec is limiting where NetApp can move too. Reference to Provision Manager exemplifies this: “policy-based automation to provision your entire NetApp SAN and NAS infrastructure from a single console.” . That goes back to make opening statement; this is all well and good in a NetApp only eco-system but its only 1/2 the IT picture.
I like NetApp, I dont want to see them wither, but they need to start thinking differently.
That means, start thinking out of the “NAS” box.