Earlier today I had one inquiry with a national government looking at the option of building a private cloud. A few hours later, out Client Support Organization – which manages client inquiries and put them into our agenda – flagged that a second national government wanted to discuss about how to build their private cloud.
In both cases I tried to understand where those decisions were coming from. In the first case, I had the client on the phone or – better – a client’s consultant, and as I went through the usual line that cloud is nothing else than a sourcing option, their attitude gradually shifted, and they seemed to understand that building their own cloud is not the only, unavoidable option.
As far as the second client, I did not have a chance to speak to them yet, but made clear that they would find value only by keeping an open mind and being ready to tell me why they had opted for a private cloud, otherwise they should rather talk to my colleague Tom Bittman, who covers private clouds.
I have found that the attitude of building a private cloud by consolidating existing data center assets is quite common both for large, insourced agencies, as well
as at a whole-of-government level. Last week, the conversation with a statewide shared services IT organization went along the same lines: how should they go for building a private cloud? No hesitation about using cloud computing for anything else, including stopping doing things they have always done.
But is building a private cloud the smartest thing to do? Does the business really need all the scalability, elasticity, pay-per-use delivery style? It is very true that government entities do need data sovereignty, and being in control of “their” cloud makes this easier. On the other hand it seems to me that this is not dissimilar from the argument in support of full outsourcing (or full insourcing). Now, who would really like to have all his eggs in a basket? How can one private cloud possibly be the best way of providing a moe rational access to IT resources?
Reality is that for different kinds of applications, security requirements, workloads, different services may be needed to get the best possible value for money. Investing on building one own’s private cloud means investing capital, skills and credibility on one single basket and, for how good that can be, this may prevent from seizing better cloud service opportunities as they become available.
In the early days of the cloud it was all about “public” cloud, with the attractive analogy with utilities like electricity. Now it is all about “private” cloud as well as “community” clouds. But this is a black & white approach: haven’t we learned after decades spent in dealing with external service providers that the best option is selective sourcing? So, why should this be an different for cloud?
Category: cloud Tags: private cloud

Andrea Di Maio





































































































3 responses so far ↓
1 Sidsavenue June 14, 2011 at 7:12 am
Security is the main concern for organizations to go on Private Cloud. I also feel apart from the obvisous benefits, cloud can help kill Piracy…Read this. http://sidsavenue.blogspot.com/2011/05/be-on-cloud-to-kill-piracy.html
2 Ross June 15, 2011 at 1:56 am
Agree on governments not owning their own private clouds Andrea. I can see a federated private cloud model satisfying the short term requirement until security in the public cloud is addressed.
3 Mark Grundy June 28, 2011 at 11:35 pm
I think that the title didn’t match the body, Andrea. A better title for your argument might be “Why should national or state governments pursue their own private clouds?” Because as I read it the thrust of your argument is that organisations should know why they’re doing it, rather than just assume they should do it.
I agree, but there are still valid reasons for doing it on both WoG and agency levels, and as the technology and methods improve, the reasons are likely to grow rather than shrink.
And does owning cloud for one purpose necessarily preclude leasing cloud for another? Would we argue the same for a vehicle fleet say?
Best regards,
Mark