CloudOps is different than traditional ITOps has been saved
CloudOps is different than traditional ITOps
CloudOps, as the name suggests, refers to cloud operations. It’s a combination of network, security, performance, device management, and help desk tasks. Basically, it covers all things needed to run a set of cloud-based business applications, ongoing.
A blog post by David Linthicum, managing director, chief cloud strategy officer, Deloitte Consulting LLP
Many people look at CloudOps as a continuation of ITOps, or the next generation of ITOps. This is an incorrect assumption when you consider the different skills and technology that are a part of both.
The surprise that most enterprises experience as they deploy their cloud-based applications is that CloudOps is something new. New in the processing and work streams, new with the technology and tools, and a new set of skills required to run a cloud-based system ongoing.
Let’s breakdown what’s new and what’s important with CloudOps:
First, you need to understand that all things will run virtually. Traditional ITOps means the ability to go into the data center and troubleshot a hardware server using traditional tools; sometimes a screwdriver and a wrench. CloudOps means that you’ll never actually see or touch a server, so you must rely on the cloud provider’s native tools to find and fix issues with virtual servers that only exist in the memory of a cluster of servers.
Second, you need to find a good set of ops management tools to abstract you away from the complexity of CloudOps. Let’s face it, we’re moving from a complex IT environment to a distributed complex IT environment. Unless you’re equipped to deal with the added complexity by abstracting yourself away from it, you’ll likely hit a “tipping point” where the number of systems you operate exceeds your capacity to track those systems during operations. Typically, service suffers, users are impacted, and you’re unable to address the issues quickly enough.
Finally, this is a proactive not reactive operation. If a user finds a problem with the cloud-based applications or data, you’ve failed. There should be an ongoing analytical process that spots issues as they occur as well as predicts issues that are likely to occur, and then corrects them before they become issues. Examples of this include moving data from one storage system to another if performance is becoming an issue on a storage instance, or blocking an IP address that seems to be attempting to attack the systems. All of these activities are automated and driven by gathering data over time and determining patterns.
CloudOps is perhaps the most overlooked aspect of cloud computing. The lack of a good CloudOps approach and technology set is the biggest reason cloud computing fails within an organization. Don’t be that organization.