Managing virtual servers and applications in the cloud is hard because it requires not only major changes to technology and process, but also a mindset shift. You need to align your tools and strategy for automation with your team’s “cloud awareness” mindset. Otherwise, moving to cloud infrastructure can end up costing you more time and money than physical infrastructure.
The need to align your methodology with your cloud awareness mindset is a big reason why there is no single, correct “recipe for success” for leveraging the cloud. The tools and processes that will work best for a given team at a given point in time depend on its actual state of cloud awareness-that is, how you view and treat your server and application instances.
As one of the most experienced boutique Boston cloud consulting firms targeting startups and SMBs, we’ve worked with clients at every point along the cloud awareness continuum, from “legacy infrastructure” (think racked servers in a data center managed by hand) to “cloud-aware” (think continuous delivery, ephemeral instances, and no panic attacks if you lose one or even ten cloud servers in an outage).
We’ve also seen the Boston DevOps scene, as well the overall industry focus on DevOps, expand considerably even in just the past year. Likewise, the pace of cloud migration around Boston and elsewhere has accelerated even more. This is true not just among startups and small tech firms, but for all businesses that rely on IT or software development as a differentiator.
Yet many teams, even though they’re “on cloud,” still have vestiges of a legacy infrastructure mindset. For example, teams at the cut/paste stage of cloud migration can successfully save on hardware costs by virtualizing servers. But they still tend to treat their virtual servers as if they were sitting in racks in a data center: they give them names, manage them individually and keep them up-and-running indefinitely. They also tend to struggle with deployment and administration bottlenecks similar to what teams working on traditional data center infrastructure frequently encounter.
In contrast, teams at the “partially cloud aware” stage are actively leveraging the cost and speed advantages of public cloud. But they probably have separate processes for deploying code and infrastructure.