A featured contribution from Leadership Perspectives, a curated forum for enterprise technology leaders, nominated by our subscribers and vetted by the CIOApplications Editorial Board.
Grinnell Mutual Reinsurance Co
The Maze-Operations journey to a DevOps outcome
The days of having a team dedicated solely to the daily operations of systems are over, and the complex series of twists and turns of a changing landscape lead to complicated choices for each organization. Individual and team responsibilities continue to evolve as development and automation lead changes in day-to-day activities and create challenges for dealing with legacy applications all while DevOps teams strive to move code faster and more efficiently.
Historically, legacy applications were developed, tested, and then turned over to an Operations team to make sure that they ran, errors were handled, and run books were kept current. If something went wrong, the operations team would follow up, engage the appropriate resources, and triage the event.
This same team was also responsible for making sure the systems stayed patched, the software was up to date, Development was initially done at a slow and controlled pace, with the path mapped out before anyone stepped foot into the production maze.
Agile development addresses these shortcomings and has become more prevalent in organizations but requires finding a new route through the maze and forcing Operations teams to streamline and to partner closely with the Development teams, planning work in smaller segments completed within sprint timeframes.
But the Agile method comes with caveats, too. Operations teams will need to thoroughly understand all the moving parts, create backout plans and disaster resilience as automated regression testing inevitably becomes part of the process. It requires vigilant monitoring of support issues and keeping alerting functions up to date. As Agile teams continue to move faster, organizations must continue investing in these capabilities. The ability to react quickly to changes, mitigate downtime, and maintain effective system monitoring require transparency among teams and an open communication path.
The final turn to a new Operations norm is taking the DevOps path. A combined DevOps team, along with more automation is changing how almost all operations are handled and allows development teams to move code, perform backouts, and create changes at a speed that unintegrated operations can’t manage.
In a DevOps team, the newest challenge is keeping Operations requirements in focus, even as the development teams try to move as quickly as possible. Finding the right balance between speed and success requires that things like security, monitoring, and metering be built into the automation from the beginning.
As DevOps teams continue to evolve, Operations will undoubtedly morph, looking nothing like they have in the past. Although, there will always be a need to have operation types of accountability for functioning systems and applications, the discussion of which teams own that accountability will continue to change.
The importance of flexibility and adaptability to change will only increase, and the integration of those who are writing code with those who are supporting and developing it will eventually erase the differences between those roles. Separation of duties may blur with each turn, but the need to replenish that control with new logging and advanced monitoring will be increasingly crucial. The important fact is to make sure that the “Ops” doesn’t get lost in the maze of “DevOps”.
