Technology continues to evolve and both technology professionals and luddites continue to evolve with it. This means that they way we use technology changes. Changes cause us to either adopt new technologies into our existing operations to take advantage of capabilities that were not there before, or restructure operations to take advantage of new opportunities and new efficiencies that were not possible before. Adopting iPhones from flip phones was an integration of new technology into their existing process. Adoption of Apple TV and Netflix caused changes in behavior. Most obviously, that there are no more trips to the video store. More subtly the way content is consumed has changed, several episodes of a series are consumed consecutively, rather than piecemeal spread out over time. Interruptions invoke the pause button rather than waiting for a commercial breaks. The previous evolution of TiVo, quickly meant that shows that couldn't be time shifted suddenly became inconvenient, commercial breaks that couldn't be fast forwarded were a nuisance. In short live TV has become a pain to watch.
The introduction of cloud has introduced a new variable into the technology equation. Businesses have a choice on where they procure services. IT infrastructure can be built out to provide services to support the needs of the business or services can be purchased for a cost from a cloud provider. The appropriate solution depends on the needs, of the business. How quickly does the business need it? How long will it be needed for? What are the costs? What information will be involved? What are the security concerns?
Cloud has the advantage of being extremely elastic. Services can be rapidly allocated and deployed, and when they are no longer useful they can jsut as quickly be disbanded. To the end user the cloud hides all of the underlying layers of complexity and replaces them with a shopping cart. Ordering IT services is now as easy as buying a book online.
The business demands to be able to roll out applications and services much more quickly. For comparison look at how software updates cycles have shrunk, to where apps on your phone are now updated weekly. IT needs to be able to quickly provision services to meet the needs of the business in real time. The model for IT infrastructure is now changing to one of pool based resources where slices of these resources can be allocated on demand, and applications can be spun up in minutes. Virtualization is a key enabler of this evolution, but better management and operations tools are critical to removing the underlying layers of complexity. With these tools it is possible to provide the business with a shopping portal for IT services, where they can immediately access these resources and spin up applications. In the date center there are pools of compute, storage and networking resources being provisioned to enable this process, but it is entirely automated. Charge back tools charge the business units for the resources they use. Fee for service discourages IT sprawl, services that are no longer being used and legacy applications are disbanded. Compared to physical systems, which remain in place consuming space and resources until they are fully depreciated and can be removed. In this model we treat IT resources as we do the gas tank in the car, we fill it up with what we need and when it starts to run low we add more. We don't stop at the gas station for every trip to the store, or commute to work.
What does this mean for IT professionals? Our own internal customers will source their needs from the provider that makes the most sense for the business. With the introduction of Cloud, IT customers now have a choice. It is critical for internal IT to be able to meet those needs of their internal customers. Growth in scale of IT is leading us to manage resources as commodity services to the business based on the SLA they provide. Stovepipes of infrastructure are being replaced with compute, storage, and networking as a service. Changes in the density of the technology are making this necessary. As we approach multiple petabytes of data we can no longer manage individual gigabytes of data. Upfront allocation methods that pre-allocate based on anticipated need are inefficient and lead to wasted resources. Implementing pools of storage that can dynamically grow or shrink depending on the needs allow for higher utilization. Hardware is evolving toward denser footprints, which require less management. Resources are being offered in pre configured blocks, or a datacenter in a box. Fueling this trend Intel is developing servers on a chip, and HP is developing a server down to the size of a disk drive and can mount them in trays similar to the ones used for disk drives. At this scale the need for IT expertise in storage, networking, and servers is as critical as ever, but we can no longer effectively micromanage each individual component. That granularity of management requires the use of tools and automation for the infrastructure to grow out cost effectively at scale, and for internal IT to be able to offer the same services and response time of the cloud.