Storagebod Rotating Header Image

No Pain, No Gain?

I always reserve my right to change my mind and I am almost at the stage that I have changed my mind on blocks/stacks or whatever you want to call them? And for non-technical and non-TCO related reasons.

I think in general componentised and commodity-based stacks make huge sense; whether you are building out private or a public infrastructure; a building block approach is the only really scalable and sustainable approach. And I wrote internal design documents detailing this approach eight or nine years ago; I know I’m not the only one and we didn’t call it cloud…we called it governance and sensible.

But where I have changed my opinions is on the pre-integrated vendor stacks; I think that they are an expensive way of achieving a standardised approach to deploying infrastructure and I have not changed from this opinion.

However I think that this cost may well be the important catalyst for change; if you can convince a CFO/CEO/CIO/CTO etc that this cost is actually an investment but to see a return on the investment that you need to re-organise and change the culture of IT, it might well worth be paying.

If you can convince them that without the cultural change, they will fail….you might have done us all a favour. If it doesn’t hurt, it probably won’t work. If it is too easy to write things off when it’s tough…it’ll be too easy to fall back into the rut.

So EMC, VCE, Cisco, IBM, NetApp, HP etc….make it eye-wateringly expensive but very compelling please. Of course, once we’ve made the hard yards, we reserve the right to go and do the infrastructure right and cheap as well.


11 Comments

  1. Or you do what Nutanix does in it’s block — true convergence brought about by software. We literally bring compute and storage together in a scale-out environment, reducing footprint, power, network complexity, and costs. There are no 3 or 4 sales people to pay along the way. And that by itself is a big capex saver!

    1. Martin Glassborow says:

      Not sure that you really understood the point I’m making; this is not really about product, this still about organisational culture. Perhaps you really need to get some serious skin in the game to ensure that the necessary cultural changes are made.

      1. Look at history, that is what they said for various things I write below. The mid-market is not religious, and the mid-market is *the* one that makes new technologies work. Organizational structures are relevant for F500.

        (a) Apple is only for the consumer; an enterprise would never use a device that does not have a Blackberry-esque security model
        (b) Social networking is a fad only good for personal communication
        (c) SaaS is for SMBs, not for F1000
        (d) Linux, Apache, and MySQL are for web companies, not for the enterprise
        (e) Digital cameras are good as embedded toys, not good enough for professional photographers
        (f) Dedup is for those who want to bring the wrath of auditors
        (g) Virtualization is a test-n-dev toy
        (h) iSCSI is for kids; grown up folks use Fiber Channel

        And the list goes on and on. See my blog post from last year: http://www.nutanix.com/blog/2011/08/16/we-are-officially-open-for-business/

      2. BTW, this whole organizational culture thing is overrated as well. Culture inertias are often used as an alibi for lack of market segmentation and thorough go-to-market strategy (http://ow.ly/aFYKf).

  2. Martin Glassborow says:

    Dheeraj,
    sorry I still struggle to see what point you are trying to make apart from pimping your product? Pimping your product is fine but without some kind of context and commentary, I’d rather you kept it on your blog.

    Best of luck,

    Martin

  3. I don’t think you like disagreements on your blog. I am trying to say that organizational cultures matter only among the laggards. Innovations are not spurred by laggards.

  4. Martin Glassborow says:

    I completely and utterly disagree, when you are dealing with large organisations which number of 1000s of people; organisational cultures matter hugely.

    Where you have siloed teams with different teams supporting operating systems, networks, storage and applications; you need to change the culture. You need to converge teams and re-structure to support a new paradigm. Just wishing it does not make it so.

    I work for a company which many people would say innovate in their space rapidly and I would be a tiny bit surprised if anyone was to call us laggards but for us to truly innovate internally, we might well need some re-alignment and this is the same for many companies of our size.

    It is like trying to turn a super-tanker. It’s not easy and you need a good reason to do so. This is my reality, yours may be different.

  5. I am not totally discounting the role inertia plays in product adoption. If convergence is true innovation, and we know it is, it has to start in the appropriate backyard… and Fortune 1000 is not that backyard. That’s all I am saying. It’s better to start selling convergence at places where virtualization and storage roll up to a Director of IT, for whom territories don’t matter. It’s one team, one mission, and no room for politics or territories.

  6. Martin Glassborow says:

    Okay, I work in a FTSE100 company…so that is my context; your context is very different.

    And is convergence true innovation? Or a return to the mainframe reality?

    1. We used to have a Motorola RAZR (cellphone), a Palm Pilot (personal organizer), a Sony Walkman (media player), and a Windows PC (word processing, web browsing).

      Then came Apple iPhone, and the rest is history. It converged 4 devices into one. It simplified user experience. It changed the way we think and live.

      Convergence is about simplification. We went too far trying to optimize the network between compute and storage. It’s time to bring them back again.

  7. Lee Johns says:

    Interesting piece Martin and I concur that organizational structure and culture is key. I spent 27 years at HP and I was responsible for the HP management infrastructure for servers and blade servers for a good part of that. That work was the instigator of alot of the current “Convergence” Fad.

    The reality is that it is easier to sell a customer what they want than it is to sell them what they need.

    Previous cultural shifts in adjacent markets are not the reason you should buy anyones product.

    I also work at a startup with Starboard Storage. We do not need a cultural shift to sell our product to our target market but I for one hope that EMC and HP et al do change corporate IT culture. It would make my life easier because more people would want what we have to sell.

Leave a Reply to Dheeraj Pandey Cancel reply

Your email address will not be published. Required fields are marked *