Traditional IT infrastructure generally comprises a range of technologies and platforms with minimal standardisation – although the level of standardisation varies heavily between organisations. This infrastructure will use a combination of virtual and physical hosting platforms, supporting long-life traditional infrastructure with allocation-based pricing and the slow adoption of new technologies. Traditional IT is often completely customised – built specifically for the organisation’s requirements and only upgraded with new features after the organisation’s demand.
Traditional IT is often accompanied by a custom-built usage philosophy – the ability to have any build of configuration supplied based on the application’s requirements – which over time leads to a bespoke and operationally complex environment. The transformation away from Traditional IT towards cloud infrastructure tends to be a greater change to the consumer’s culture and organisational processes rather than the underlying technology.
One of the key points to stress here however is the comment about “long-life” infrastructure. Traditional IT is typically used to host an application that may have a life time of multiple years before the operating systems upon which it sits is refreshed. It would still (hopefully) be patched during that time, but once the OS was deployed, it typically hung around for a while. This means it was also asset managed, described in a configuration database, probably paid for by the month, and with many tools designed to keep it running, and report when there’s an operational anomaly.
One of the interesting decisions about moving to the cloud, is how this kind of long-life application would be supported. Either the application is replaced with one that has a much shorter life cycle – and thus provisioning new infrastructure with each new release of the application, or the traditional tenants of hosting long-life applications needs to be adhered to in the cloud environment.