This post was written and contributed by Sue Poremba, a freelance writer focusing primarily on security and technology issues. Sue blogs via Contently.com.
Application developers and IT operations managers have historically been in separate camps within enterprise. What could bring the two entities together? The DevOps movement.
DevOps envisions a world where these two groups are collaborative, agile and high functioning to increase efficiency and reduce the production risk associated with frequent changes. As the role of development and operations changes with the advent of virtualization and cloud computing, DevOps is at the forefront of almost every IT strategy.
“IT is already struggling to keep pace with demands for efficiency, scale and reliability across an increasingly virtualized environment, and pressure is mounting to do more thanks to the availability of public cloud services,” says Brett Adam, chief technology officer of rPath, a Rackspace Cloud Tools partner.
“Dev” and “Ops” teams each face the same broad issues, yet because they work in silos, they are frequently working at odds, resulting in application downtime, lost revenue and missed opportunity, Adam points out.
Fixing the Dysfunction
Business unit leaders are becoming savvy to this dysfunction, and are starting to look elsewhere to achieve their goals, leaving IT with a difficult choice — join forces to improve performance, with all the complexities this entails, or risk losing the organization and jobs to the public cloud.
For some greenfield large-scale application projects and new cloud deployments, organizations are starting off with DevOps-style operations, with strong collaboration between the Dev and Ops teams. But for the majority of IT organizations, they are just coming around to realize the need for coordination. There are distant silos even within the Ops and Dev teams, so they have a long way to go, Adam explains. In organizations where the Dev team is in charge of deployment (especially of large Windows applications), they are implementing operations enhancements into the development process. However in organizations where the Ops team is in charge of deployment, there aren’t as many big changes in the Dev process.
However, they aren’t combining their efforts just yet; instead, they are preserving their current organizational structure. That’s why Adam says breaking down communication barriers to fix the dysfunction is so important. And that’s where the cloud comes in.
“The very nature of moving to the cloud serves as a forcing function for addressing some of these process challenges, though it might come under the guise of “private cloud” vs. “DevOps,” Adam explains.
The economics of cloud computing are extremely compelling, and enterprise developers and lines of business are looking for ways to reap these benefits from a time and cost perspective, leaving IT operations out in the cold.
For organizations that prefer the “control” afforded in private or hybrid cloud computing models, these initiatives are bringing into focus the need to standardize, automate and accelerate application development, deployment and change. This is forcing enterprises to think differently about their application delivery models – from Dev through Ops.
The ease of providing virtualization and cloud resources with minimal IT management has companies rethinking — and shifting — their focus to IT operations, David Link, chief executive at ScienceLogic says.
“These new computing environments enable rapid deployments, exposing development and operational inefficiencies underscoring the need for IT operations to provide critical oversight and support, including automation of monitoring and management after provisioning occurs.”
For most enterprises, the IT operations function is the interface between IT and business, providing essential visibility and support, which are increasingly important and challenging as services move into the cloud.
As DevOps continues to play a more important role in IT strategy, the ability to monitor and manage cloud resources, along with physical and virtual ones as one entity, is essential in delivering optimal service delivery to businesses. This includes incorporating the right cloud management tools and processes to strike a balance between development and operations.
Fundamentally, the business needs IT to deliver applications; infrastructure is just an enabler for applications, Adam says.
“When Ops applies the automation lessons of DevOps, IT can finally start keeping up with business demand for applications, delivering them with consistency, reliability and speed.”
Questions for discussion:
How do you define DevOps?
Are you running a DevOps team or do you separate Dev from Ops?
Has the cloud made DevOps run more smoothly?
Are you an application-centric or an infrastructure-centric organization?