Did these "people involved in cloud computing" represent companies that depend on revenue and profits from current current client/server applications and platforms? Many companies have entered Cloud Computing so they are not left behind, even though Cloud Computing represents a huge risk to their current cash cow client/server applications and platforms. These companies (including several mentioned in this report)have publicity stated that profit margins from Cloud Computing are much lower than selling licenses. Despite having their own cloud offerings these companies benefit financially from trying to slow Cloud Adoption and it provides them the time needed to try and catchup on the technology.

As for their comment, "Cloud computing is—at least today—inappropriate for enterprises" I would point to salesforce.com with its growing list of Enterprise customers including Dell, Cisco, Aon and others with tens of thousands of users. These customers are leveraging salesforce.com's CRM applications and platform today.

People are beginning to use external cloud services where this issue with migrating existing apps isn't a problem. (So, for new apps.) It's also one of the reasons that we've heard for people investigating internal or private clouds. But even there, the key will need to be creating those "cloudy" benefits (elasticity, sharing of resources, etc.) using the data center infrastructure that someone already has in place. More pro/cons re: internal clouds here: http://datacenterdialog.blogspot.com/2009/01/are-internal-clouds-bogus.html

Interesting post.....however "Cloud computing is—at least today—inappropriate for enterprises" seems like an odd premise. Most Enterprise Class companies have 1000's of applications. The suggestion that none of these applications should considered The Cloud as an appropriate deployment platform.... seem's a bit narrow minded.

There are many enterprise applications that do not require the qualifications mentioned in your blog. For example, the Cloud has proven a very attractive platform for Collaborative, Event Driven, and most Composite applications....and an ideal platform as a virtual Test Lab (Cloud Testing). At the very least....choosing a deployment option for your web applications should be an application to application decision.

Good article Bernard, but I believe the issue is not so much with the migration of the applications as with the migration of mission critical data - see my own blog post at http://www.celona.com/newsevents/blog/2567

Tony Sceales,
CTO Celona Technologies Limited.

The Cloud is not ready for the Enterprise. Will simple public services may be shifted there, critical core services managed by regulation cannot be delivered from the Cloud yet.

We are actively researching the feasability of implementing Cloud services to ourselves and our clients. One of the greatest deal breakers at this time is the requirement for privatized networks. Our clients value their data and must insure it is transported in a controlled and effecient manner. Until the Cloud implements MPLS, leased line, or some other WAN technology, it will remain a web host solution.

Hi Bernard,

You raise some interesting points, I especially like the Sinclair Lewis quote in the post you made to the google group.

I've heard many of these arguments before and they are valid and true, but becoming less valid over time.

For example, we had a Sr. Purchasing Director from one of the top 3 auto makers sign up to use our cloud service because we offer the ability to run applications on the same Platform in the cloud or behind the firewall.

In just a few weeks -- by herself, with some help from our support staff -- she migrated a legacy peer review system to our platform. She can easily synchronize this system from a our cloud hosting environment to her internally running platfrom on corporate managed infrastructure with a click of a button!

When applications can move seamlessly between the cloud and internal servers like this, there are fewer reasons to argue against the cloud. If, as it turns out many times, the systems developed outside the firewall must be moved inside the firewall, this is super easy. If the opposite is true and you build on our platform internally, but want to leverage the cost savings and scalability of a cloud environment at some point in the future (when your CIO wakes up), simply deploy the application to our cloud environment and you are done -- in one click (plus typing in a username/password and servername to deploy to).

Another customer, a large insurance company, has used our enterprise platform to dramatically increase their Time-to-Market. Prior to using our platform, their IT department typically took 6 months to deliver a software prototype to meet a particular customer need -- just a prototype!

Within just months of purchasing our platform, one person in their organization had prototyped three entire separate and quite complicated applications managing information across multiple states and offices around the United States.

Now, as you say in your post, there are existing applications that need to be ported. To do that with our platform, all the user needs to do is export the existing data to excel or .csv files and then import them. Our platform does most of the of the work from there, introspecting the relational data model to support the data automatically! When I say "most" I mean, UI navigation and data security still must be configured to control access to information and functions within the application.

So, Sinclair Lewis may have been right, but I think in the end, it will be the CTOs and CIOs who "get it" that will retain their authority.

If your goal in life as a C(I/T)O is to have a bunch of people reporting to you and controlling the spending of "too many dollars" then eventually, the CFO is going to be asking why their competitors are reducing their IT spending by outrageous sums while you neglect the reality of the world around you for some personal desire for power.

In an organization, Knowledge is power, information is power, efficiency is power, and money is power.

Executives who make decisions optimizing the use and management of those things are going to win and those who let personal issues "cloud" their judgement and refuse to cloud their Information Systems are going to lose.

It's as simple as that. The only question c-level execs need to ask themselves is, "Do I want to have more experience in the cloud than my competitors when they wake up?"

a primary reason for the cloud is to move from the dinosaur app now operational... as was the case for GE Supply Chain... Amazon seems to have their act together...

Cloud computing is a design approach that promises significant business value. Like any other disruptive business model from the past, there are technical, legal and human challenges that must be surmounted to make this a reality. None of these challenges requires a violation of the laws of physics. They will be overcome and the companies that do so will be the winners.

We have built our business (http://www.practicechoice.net) on this need to migrate legacy applications, in our case software for medical clinics and practices, and can tell you from experience that it is possible and is being done today.

Certainly our scale doesn't match that of the largest Fortune enterprises, but an iterative adoption strategy that picks the right tool for the right job will certainly meet with success.

As as former CIO of a large company, I have a more optimistic point of view, although many of your comments are right on target.
I believe that there is indeed a cost advantage, once the migration issue is resolved (this is definitely hard). I developed my own analysis in http://informationsystemsbiology.blogspot.com/2008/11/very-cloudy-weather-will-it-rain-on.html
I believe that there is a "technical issue" related to performance requirement (cf. your point about SLA) which is more subtle than it looks. I write "SOA is not scale-free" to represent the fact that not all services are created equal as far as "cloud migration" is concerned.
However, once the hype is over, there are many opportunities for cloud-based corporarate IT.