First published on the Rackspace blog.
A lot of discussion round the traps has occurred around the fork lifting of legacy applications onto the Cloud. Recently I interview Christian Reilly about this, and other issues – his feeling was that the reality is that there simply isn’t the budget within organizations with legacy applications to spend the time to forklift those apps onto the Cloud – it is for this reason that Reilly was a little negative about the cloudy opportunities for organizations with a host of legacy applications.
It’s worth watching my video with Reilly for a glimpse into the reality (or at least one particular reality) of life in enterprise IT.
It was interesting then to watch a presentation given my Adrian Cockcroft, the oft-lauded Cloud Architect at Netflix and perhaps the person who has done the most to further the interests of the Public Cloud from a customer perspective. In his presentation, Cockcroft reflected on the fact that “fork-lifted” applications, i.e. those that are taken in situ from on-premise infrastructure and moved into the Cloud, just simply don’t work well. Cockcroft contends that these applications prove too fragile and have to many datacenter oriented assumptions to perform effectively. As Cockcroft says; “in the datacenter, robust code is best practice. In the Cloud, it’s essential”.
Instead Cockcroft advises organizations to port to Cloud architecture saying that it needs a short term investment but delivers long term paybacks. It’s really interesting reading this in light of my previous post about the 80/20 rule – in that post we looked at the realities of IT departments, with 80% of budget going to simply “keeping the lights on” leaving very little to pend on green fields projects.
It’s one of the conundrums of Cloud Computing, as we looked at in our Cloudonomics paper, Cloud has the ability to save money on a “like for like” basis for an organization. But for those organizations burdened by legacy applications, sometimes it takes an investment in order to realize those longer term savings. The benefit that people like Cockcroft have is that they work for an organization without the need to consider legacy approaches – for them the decision is a simple one of significant ongoing CapEx to build out infrastructure, versus OpEx for utility based Cloud services.
The reality is that this won’t change any time soon – for the foreseeable future there will be existing organizations that look at the realities of moving existing applications onto the Cloud – for many of these organizations, and despite the vehemence of evangelist like Cockcroft, there simply won’t be sufficient budget to move these workloads. For these organizations it will instead be a combinations of different things moving to the Cloud;
- New workloads, test/dev etc
- Generic workloads (email for example)
We’re at a paradigm shift in IT – paradigm shifts always include a degree of pain, and occur over a protracted time period. That doesn’t lessen the importance of the Cloud any, just brings a does of realism to the discussion.
As always, I’m keen to hear your thoughts, oh and we’d love you to join in some of our webinars or read the whitepapers –the CloudU homepage is here – and you can register to have updates sent to your inbox (in a non-spammy way of course!) there.