Blogs

RedMonk

Skip to content

Do Private Clouds Mean Stormy Weather’s Ahead for the Incumbents?

Asking whether private clouds will find fertile ground within the enterprise is asking the wrong question. Who doesn’t want a more elastic infrastructure?

The drivers for internal cloud infrastructure will be numerous. From necessary modernization to NIH to security and compliance concerns, the justification for private cloud investment will vary, but the end result will be the same. The question is will the suppliers be the same as usual, or will it be an an entirely new crop of contenders? Barron’s Mark Veverka seems inclined to argue the latter. His recent piece “A Private Party” uniquely positions the private cloud as a clear and present danger to traditional systems suppliers like Cisco, Dell, HP, IBM and Oracle. Uniquely because while many would argue that the public cloud is a potential threat to the incumbents, the private cloud is more often than not perceived as opportunity.

To his credit, Veverka does acknowledge that the short term prospects with respect to the private cloud are positive:

In the very near term, companies will continue to invest in their own private cloud-computer systems. That will benefit the traditional tech behemoths that sell servers, storage, personal computers and business software, such as IBM (IBM); HP; Dell; Oracle and Cisco.

Left unsaid is how the longer term opportunities will evaporate.

It is true that to date, the large systems vendors have shown little inclination to invest directly in product around the cloud software opportunity. Microsoft of course has Azure, but that is a public cloud investment first and a private cloud opportunity second. None of Cisco, HP, or IBM currently offer a SKU for a cloud platform [Update: IBM points in the comments below that they have, in fact, two SKUs for appliance and software based cloud platforms]; Dell, for its part, has partnered with Joyent. More than a few of the vendors would argue that this is the logical strategy in light of the legal uncertainty around the IP of Amazon’s published interfaces (an attractive customer feature). Big vendors instead are tactically outsourcing the risk to smaller – and therefore less attractive from a litigation perspective – third parties. But the larger vendors software holes notwithstanding – which, as the Barron’s piece speculates are likely to eventually be plugged via acquisition – it is difficult to imagine private clouds becoming anything but accretive opportunities for large systems providers.

Few enterprises are likely to follow in Google’s footsteps and assemble their hardware and storage from whitebox components; they’ll simply buy it from their existing suppliers the way they always have. And while cloud infrastructure suppliers such as Cloud.com, Eucalyptus, GoGrid, Joyent and RightScale are indeed enjoying varying degrees of success in selling into the enterprise, a portion of that market will be reluctant to trust a smaller supplier with a strategic role in their infrastructure, which a private cloud must assume by design. Which again opens the doors for the traditional vendors. The opportunities for existing suppliers in the private cloud, then, seem quite extensive from the top of the stack to the bottom. Which explains why they are embarrassingly eager to drive customer cloud discussions towards private cloud offerings: that is where the opportunity lies, not the threat.

When the threat arrives, it is far more likely to emerge from the public cloud. Besides the economies of scale that Amazon, Google and others may bring to bear, which might be matched by larger commercial institutions, the experience of web native entities operating infrastructure at scale is likely to prove differentiating. It’s not just a matter of running a datacenter at scale; it’s knowing how to run applications at scale, which means understanding which applications can run at scale and which cannot. The cloud – public or private – must be more than a large virtualized infrastructure. Whether or not private cloud suppliers and their customers realize that yet is debateable; whether Amazon, Google et al do is not.

Were I to look for threats to the incumbents – the kind of threats that could put “enterprise-technology vendors…at risk of becoming obsolete” – I’d look to the public cloud [coverage], not the private. They say danger and opportunity are two sides of the same coin, but I know which side I’d bet on if we’re talking about the private cloud.

Disclosure: Cloud.com, Dell, Joyent, HP, IBM and Microsoft are RedMonk customers. Amazon, Google, GoGrid, Oracle RightScale are not.

Categories: Cloud.

Tags: , , , , , , , , , , ,

  • http://achurchassociates.com/ Ric

    I’m puzzled … where are the cost and elasticity benefits for enterprises if they own their own cloud? If they have elastic compute capability, then what they have is no more than an over-engineered datacentre with excess capacity, surely?
    I understand the benefits of virtualisation, etc, but it seems to me that “private cloud” is an oxymoron …

  • http://redmonk.com/sogrady sogrady

    The benefits are likely to arrive, at least in part, from more efficient utilization of existing resources. In other words, the age old promise of enterprise IT technologies: no more 5-15% utilization rates.

    It’s true that in house hardware mitigates to some degree the elasticity advantages – specifically OPEX and CAPEX minimization – but that’s assuming that private clouds are not integrated with public resources. Many private cloud offerings, in fact, include relatively extensive provisioning for marshalling and deployment of public cloud resources. See Eucalyptus with its EC2 API.

    Private cloud is about bringing some of the public cloud benefits inside the firewall, then calling out to external resources for extreme elasticity.

  • http://achurchassociates.com/ Ric

    I guess my point is that the internal “cloud” is little or no different to what enterprises were well on their way to doing anyway. I buy all the efficiency arguments for virtualisation, etc., (my previous employer rented space, owned the servers, and (18 mths ago) had achieved close to 100% virtualisation of the hardware), but I don’t buy that as “cloud computing” … semantics, perhaps?
    Now – “overflowing” out into the public infrastructure (e.g. Amazon, Rackspace) I can call “cloud”, and I was arguing that for seasonal peaks in CPU requirements and develop/test environments; I suspect I was a little early though :)

  • http://achurchassociates.com/ Ric

    PS – I think at the time I was suggesting CohesiveFT’s VPNCubed for management of a hybrid environment (http://www.cohesiveft.com/vpncubed/) – Eucalyptus is a little more recent (and more open?).

  • Ric Telford, VP, IBM Cloud Services

    It’s hard to understand what Veverka is trying to say. I read it as him saying that private clouds are accelerating the transition to public clouds, as he uses the phrase “…next step is the outsourcing of many data-processing services.” He then goes on to say this is a threat to folks like IBM. Well, who is the biggest provider of outsourced data-processing services? Hmm. Cloud is really just an evolved delivery model for IT services and something we strongly embrace in IBM. As for your article, Stephen, I agree with a lot of what you say, except the part about IBM not having a SKU for a cloud platform. I invite you to check out http://bit.ly/aUEpdh . IBM Cloudburst is an integrated hardware/software cloud platform, IBM Service Delivery Manager is a virtualized software cloud platform. They are used for quickly deploying a private cloud, and share the same DNA as the platform we use in the IBM public cloud.

  • http://redmonk.com/sogrady sogrady

    @Ric Telford: my suspicion is that Veverka was the victim of some heavy handed editing, because ultimately the article provides little argument to support the titular claim. Indeed, the outsourcing of data-processing services he refers to actually points more to a threat from the public cloud than the private versions he suggests you should fear.

    As for the SKU argument, thanks for the reminder: point taken. My only objection would be that in my briefings on both, they lack a few features we might reasonably expect to see on private cloud platforms: notably, Amazon API compatibility, for the reasons mentioned above. But I’ve updated the piece above to reflect your comments.

    In any event, we’re in agreement, I think, that the opportunities in private cloud for IBM and other large suppliers are substantial. Now it’s a matter of execution.