In the age of automation and the shift left movement, the exact role of IT operations teams might seem fuzzy -- but it's far from obsolete.
The term shift left in DevOps means that developers address code issues during the design and development phases, rather than leave them for the operations team to resolve. This approach should be common sense -- and it should have emerged before DevOps. The financial effect of a coding problem in production will always be higher than the cost of a delay in development, and the operations team shouldn't have to pick up the pieces of poor development practices.
However, the shift left strategy has expanded, as organizations complete more tasks outside of operations. For example, developers should manage both development and testing through continuous development and delivery. And help desk loops should more closely involve business leaders in decision-making processes around the priority of issues and requests for extra functionality.
Concurrently, IT sees a shift right movement: As the use of public cloud increases, admins that work internally at a cloud provider carry out actions normally within the remit of enterprise operations teams.
How ops stays relevant during these shifts
To shift left in DevOps, development teams take responsibility to ensure that code is of sufficient quality to not affect operations. Automation ensures things run as expected and that code gets packaged and provisioned without human intervention. Meanwhile, other teams of technical professionals manage third-party platforms. So, are IT organizations left to ask, "Will the last operations admin out the door please turn off the lights?"
Of course not.
Developers will still have code that misbehaves in production. This misbehavior is not generally due to oversight but to the domino effect of changes across the platform.
Automation is great, but not always reliable. There must be a team available to deal with anomalies when the automation engine needs indication on the correct approach.
The hybrid cloud is also highly useful -- but someone needs to negotiate a deal between the organization and the cloud provider, and also manage interactions at both a human and technical level. As more organizations evolve their public cloud strategies to include multiple cloud providers, operations teams must also properly implement and manage integration. This ensures everything continues to work as it should, even as cloud providers make changes to their platforms and functional services.
Operations' role in the IT organization must change after a shift left in DevOps. Gone are the days of relatively simple systems management consoles with red, amber and green traffic light signals. Instead, IT organizations use aggregated systems that monitor the health of a complex mix of owned and third-party platforms with potentially wildly different volumes of available information. With demand to move workloads across hybrid platforms, these systems must grow more dynamic. Workload management and task automation and orchestration are not only tools that help operations -- they're tools that define how competitive organizations will be in the market. Ops teams should examine how systems such as Kubernetes, Electric Cloud ElectricFlow, HashiCorp Terraform or Stonebranch Universal Automation Center fit into a shift left strategy.
They should also work to gain a deeper understanding of the technical concepts and capabilities of various hybrid cloud platforms, as well as the financial and competitive steps their organization must take to succeed. Only through these insights can operations teams implement and enact the right policies, and thereby automate procedures, to fully support the organization and its needs.
The shift left in DevOps should not be seen as threats to the operations team. If done correctly, such approaches will eliminate drudge tasks that take up operations staff's time. They will also free up funds for true IT investment that support the business, rather than for fire-fighting that adds no real value.