Without a clearer understanding of all those bewildering acronyms will we ever get a handle on the real benefits of the Cloud?
A lot of software suppliers associate their existing product offerings with the Cloud, or as Software-as-a-Service (SaaS). But is this reaching for the sky or merely clutching at straws? Providing a hosted version of an existing client/server application and branding it as a Cloud-based or SaaS solution isn’t what I understand Cloud computing to be.
This Hosting-equals-Cloud approach is not what SaaS stands for. Rather it is the new version of Application Service Provision (ASP) by leveraging the Infrastructure-as-a-Service (IaaS) cloud offering. As a result a range of re-modelled applications, that have precious little fit with their new surroundings, are getting in the way of Cloud’s many benefits.
So it’s vital to define infrastructure, platform and software as separate services to fully understand Cloud and select the right system for you.
Investigating IaaS for Hosting
Business critical, client/server software systems may be required by only a few on a daily basis. These systems are candidates for IaaS hosting in a safely distributed mode, but are not suited to a truly evolved Cloud model because IaaS hosting won’t engage potential users who can’t afford an effective permanent subscription fee. Most people agree that such licensing costs more in the long-term than traditional licensing models – without all the economies of multi-tenanted scale that makes pay-as-you-go models economically viable.
As Michelle Price said in her recent Wall Street Journal article Pinning Down the Cloud “Despite its growing importance … many companies are struggling to pin down exactly how this technological miracle can truly benefit their balance sheets”. This is as true for the many users of applications ‘on the Cloud’ as it is for software vendors and investors. Promoting a Cloud-associated solution is fairly easy in a Cloud-supportive market, but making it a commercial success for vendors by scale up the user space is a very different story.
Elastic Cloud service provision, offered by large infrastructure companies or content provider can revolutionise the IT development environment. But are they only the domain of dominant providers? Price’s example of Google’s “no software” laptop, with applications coming directly off the Cloud is a real example of innovation. However, only a few have the technology and capacity to deliver it.
Small, innovative vendors can adopt the same principle by providing their computational functions as focused ‘services’ for a niche PaaS environment that’s too small for major markets. The key to success is to develop a rich set of functions that cover a wide range of needs for a specific domain, plus the building blocks and glue technology to allow developers to create their own individual applications by composing the building blocks.
Workflow technology fits well for this purpose. Success here lies in how to link the two communities: the developer community who can use the PaaS platform to build software; and the user community who can use the software developed in a SaaS fashion with a friendly web interface. When we developed InforSense workflow technology, we had our unique design of automated deployment of a workflow into Web application. So as far back as 2007 we already had the Cloud in mind. Today, this feature remains unique in the workflow world. When I was at the IEEE Cloud conference in Washington earlier this month, I learned that this feature is now known as the key to a workflow-based PaaS. Glad that we made one more thing right again!
I believe that the debate around Cloud is becoming increasingly muddled. Applications that harness Cloud’s true power need to be seen separately from the Cloud-claimers. The trick is to better understand the differences – and independent uses of – IaaS, PaaS and SaaS and evaluate them accordingly.