Ever since I got my first job in IT in the mid-1990’s, everyone has used a cloud in some form. Whether they referred to it as outsourcing, virtualization, central IT, or in some other way, the cloud existed and grew but it did little to stem the adoption of distributed computing. Yet at some point over the past few years, the parallel growth of these two technologies stopped and the cloud forged ahead. This shift indicates that companies have now fully embraced the cloud but remain unclear about how best and how soon to transition their IT infrastructure to the cloud and then manage it once it is there.
One of my first jobs in IT was as a system administrator at a police department in Kansas. During my time there, I was intimately involved in a project that involved setting up a cloud that enabled it along with other police departments throughout the state to communicate with state agencies. Setting this cloud up would enable our department along with others to run background checks as well as submit daily crime reports. While we did not at that refer to this statewide network as a cloud, it did provide a means to send and receive data and centralize store it.
However, the data that the police department sent, received, and stored with various state agencies represented only a fraction of the total data that the department generated and used daily. There were also photos, files, Excel spreadsheets, accident and incident reports, and many other types of data that officers and civilians in the police department needed and used to perform their daily duties. Since the state agencies did not need this data it was up to the police department to manage and house it.
This example is a microcosm of what happened everywhere. Private and public organizations would choose to store some data locally and only store certain data with cloud providers which, in the police department’s case, were the systems provided by the various state agencies.
The big change that has occurred this decade and particularly over the past two years is that the need to host any applications or data on-premise has essentially vanished. This change has freed organizations of all sizes to fully embrace the cloud by hosting most if not all internal application processing and data storage with cloud providers.
Technology largely exists at the application, compute, network, operating system, security and storage layers that make it more cost-effective and efficient to host all applications and data with cloud providers rather than trying to continue to host it on premise. Further, the plethora of powerful endpoint mobile devices that are available as phones, desktops, tablets, and/or laptops along with ever larger network pipes make it easier than ever to access and manipulate centrally stored data anywhere at any time.
Organizations must accept … and probably largely have … that the technologies exist in the cloud to support even their most demanding applications. Further, these technologies are often more mature, cost-effective, and efficient than what they possess in-house.
The challenges before them are to now identify and execute upon the following:
- Identify the right cloud provider or providers for them
- Securely and successfully migrate their existing applications and data to the cloud
- Manage their applications and data once hosted in the cloud
These objectives represent a fundamental shift in how organizations think and make decisions about their applications and data, and the IT infrastructure that supports them. This “cloud-first” view means that organizations must assume all new applications and data will end up in whole or in part in the cloud either initially or over time. As such, the new questions they must ask and answer are:
- How soon should their applications and data end up in the cloud?
- How much of their data should they put in the cloud versus retaining a copy onsite?
- If they choose not to put an application or data in the cloud, why not?
Organizations have officially embraced the cloud and what it offers as evidenced by the “cloud-first” policies that many have implemented that require them to deploy all new applications and data with cloud providers. However, migrating existing applications and data to the public, private, or hybrid clouds and then successfully managing all migrated and applications and data in the cloud, as well as determining when to bring cloud-based applications and data out of the clouds, becomes more complicated.
Helping organizations understand these challenges and make the right choices will become a point of emphasis in DCIG’s blogs, research, and publications going forward to help organizations successfully migrate their data to the cloud, and then have a good experience once they get there.