The cloud has been the swiss army knife of IT departments for over a decade. The cloud is secure, extensible and cost-aligned. Managing the cloud is easy, you just move the slider over for more or uncheck the box to delete it. The cloud changed how IT works without actually changing anything at all.
15 years ago, you would hear senior IT people talk about capacity and redundancy. Today you hear those same people talk about containers and zone resiliency. Years ago, discussions would be had about uptime and ROI. Today discussions are had regarding service health and cost optimization. I think you see my point.
The cloud has made it easier to manage all the ‘stuff’ required to make technology work. But the underlying best practices are the same. It comes down to capacity planning, skills, governance and management. When server virtualization first came in vogue, every IT Director embraced it with the idea it would “do away” with most server management tasks. What it really did was shift those management tasks into a different skillset and allow decisions to be implemented faster. Server virtualization did not remove the need for strategy and governance, it merely reduced the time and complexity for execution.
As cloud based systems take over for traditionally on-premise applications the knee jerk reaction is to assume this will reduce the required management, strategy and governance. But, much like server virtualization these tasks are not going away, the skillsets required will pivot away from Systems Engineers towards Business Analysts.
Don’t make the same mistake that IT Directors made in the 2000s. Cloud sprawl will be just as painful to fix as server sprawl was with Virtual Servers. Take these suggestions:
- A strategy of “cloud first” is fine — but make sure you take your legacy infrastructure policies and procedures with you. Adjust as required but don’t assume a reduction in management. Instead assume a shift in skillsets. Invest in training your people!
- ROI and TCO are for capital expenses. Cloud systems allow you to be much more granular in your cost reporting. Itemize and chargeback cloud instances on a per project basis. This will allow your business units to better understand the value these services bring relative to specific initiatives.
- Manage your capacity in the same exact way you did when you had most of your infrastructure on-premise. Don’t conflate the ease of spinning up a new instance with a reduction in capacity oversight and reporting.
- Resist the urge to fragment your systems! When deciding on a new vendor always weigh heavy on systems with native integration points. Fragmentation is bound to occur, but let’s avoid it for as long as possible.
- Always make sure you have someone focused on identifying risks when making cloud decisions. Historically the risks associated with migrating IT systems were mostly with successfully implementing them. Now the bigger risks are adoption, integration and security. Engineers love playing devil’s advocate — let them.
You should move core systems to the cloud, the data shows uptime is increased, costs are reduced and user approval is improved. Just don’t assume all the great IT habits your organization has created over the years are no longer needed. Keep the discipline you have built for your on-premise systems and adjust, don’t reduce, your governance and management plans.