DevOps is about bridging the often mutually competitive groups, which have conflicting goals, within an IT organization – developments and IT operations. Whereas the development group aims at implementing functional business requirements on the application side, the operations group is focused on non-functional requirements like performance, stability and availability. Further, the development group is interested in quick releases of more and more functionalities; the operations group, on the other hand, prefers fewer releases in order to minimize disruption in operations. The two groups often fail to appreciate each other’s viewpoints with the business emerging as the ultimate sufferer!
DevOps, a new movement among the IT communities, involves both operations and development to address the problem, as explained above. Being a new concept, it is yet to have a concrete definition as it lacks consensus. To some, DevOps is the streamlining of the operational part of the application release and deployment process. To others, it is the application of Agile methodology to all IT processes. Yet another group considers it an automation of IT services, especially the programmability of IaaS (Infrastructure as a Service) Cloud- based infrastructure. As a matter of fact, DevOps encompasses all these connotations, and goes beyond them.
Some of the driving factors of DevOps’s success are:
- Integration of people, processes and tools related to both development and operations groups.
- Early involvement of operations in the development process. During Requirements Management phase especially, involvement of IT curbs unreasonable demands on the operations during release and deployment.
- Integration of various tools aids in the creation of a tool chain which may cross the boundaries of operations and development. For example, a Service Ticket generating an Incident, categorized as an Application bug, automatically generates a defect in the bug tracking tool used by the development group. Once fixed, it traces back the path to the Incident as ‘resolved’ and then to closure of the Service Ticket. The whole process can be automated with integrated processes.
- Use of the Agile methodology for various ITIL process areas, focusing on small frequent releases rather than big bangs.
- Automation of as many steps as possible by using internal or external Cloud. A good example of DevOps will be to create a test server with an image of pre-installed test automation and other tools; run a complete suite of regression tests- if all pass, dissolve the server – all automated steps!
Kovair with its ALM Studio and Omnibus Integration Middleware is uniquely positioned to help customers to achieve DevOps from a very early stage. Whereas the ALM Studio is focused on development group, the Omnibus on integration of various third-party tools; together they can create the ultimate bridge between development and operations which has been missing for a long time.