The future of IT operations is critical to the future of computing. But the terms used to describe this part of the industry continue to change. While the evolution of terms is to be expected, who defines the terms for us?
Traditional IT Operations
Traditionally, IT operations has existed in on-premises datacenters. These were the women and men responsible for the infrastructure supporting every application an organization required to operate. They built sandboxes to test infrastructure and applications. When ready, these infrastructures were pushed to production. Finally, this team managed backing up data and were responsible for restoring it if something happened. They also managed the security of the entire datacenter.
In small IT shops, these IT Operations folks did it all. In larger organizations they specialized in storage, Operating systems (Linux, UNIX, Windows), virtualization, networking, or security. Common titles are IT Operations Engineer, or Systems Administrator. People with specialized skills hold titles such as Linux, Storage, or Security Engineer.
IT Operations in the Cloud
Developers flocked to build applications in the public clouds. In the cloud, they were not bound by restrictions required by traditional IT Operations people. The cloud provider took over the responsibility for keeping up the infrastructure stack. In the case of SaaS applications, such as Microsoft 365 (formerly known as Office 365), the provider managed everything from the application down. Developers no longer had to wait on IT Operations to build what was needed! They simply provided the cloud provider infrastructure requirements and a credit card. The public cloud providers made it easy and fast to get coding!
People quickly realized that to run an application at enterprise scale, they needed people who understood IT operations to help manage the infrastructure. Titles such as DevOps engineers and SREs (site reliability engineers) emerged.
Ops Should Be Boring
[being responsible for] “the software infrastructure that all of engineering relies on.” Your operation system, storage systems, distributed compute frameworks, and software development tools may all fall into this bucket.
Doesn’t this sound familiar? The concept of the post is if you want to enable smart engineers to “work towards a vision and build something great”, someone must build and maintain the underlying structure. And that operations work should at least look boring.
Buyer Beware: Sometimes Technology Labels Are Marketing
We have all heard the terms:
Cloud, Private Cloud, Serverless, Edge, Hybrid Cloud, Multi-Cloud.
I’m sure I left one out. Perhaps you’ve debated the meaning of these words with friends and colleagues. Modern container-based applications that are connected via APIs are just beginning to be more widely adopted. These types of apps require architecture flexibility such as the one that underpins the public clouds.
While words are important, vendor marketers will use descriptions to fit into an analyst quadrant or to game SEO. It’s important to understand a vendor’s claims, but not get too tied up into the boxes that they try to reinforce. What we are witnessing now is the evolution of computing. Much client-server architecture surpassed the mainframe in popularity, container-based applications that run on distributed architectures will eventually usurp client-server architectures.
The future is already here for IT operations. The words defining our field are starting not to matter. We know we are leaving one phase of computing and entering another. There will always be a need for a team that seems completely boring on the surface. But we know that team is the one that must know every aspect of every application required by our organization. The goal is to be boring, no matter what words define us.
Let’s check in next year to see how many of these terms are still on the buzzword bingo card!