Software Delivery Management Facilitates DevOps Culture Change

Culture is the biggest deterrent to the successful implementation of DevOps.



I had the opportunity to speak with Moritz Plassnig, SVP and GM SDM and SDA Cloud at CloudBees after attending CloudBees Connect. The message to enterprises in the keynote was to “embrace don’t replace” by retrofitting their current software development practices over time. Embrace new technologies while still driving critical business needs with older apps and platforms while maintaining automation, visibility, and governance.


Enterprises today have similar problems developing and delivering software to meet consumer and business needs and remain competitive: 1) drive innovation quickly, 2) reduce risk, and 3) optimize development and delivery by removing manual processes. 


Complexity is a reality. The DevOps stack continues to grow. As the adoption of DevOps grows, the process, tools, and risks multiply. To succeed, enterprises need to embrace the complexity and learn how to manage it with tools, teams, and processes aligned around a common outcome.


The right software delivery management (SDM) solution enables developers to continue using the tools with which they are most familiar and comfortable while automating delivery. Enterprises get control over common data and processes, visibility, and collaboration which results in greater efficiency and fewer errors. 


DevOps Research and Assessment (DORA) 2019 State of DevOps Report reiterates the success of elite DevOps performers:

  1. 208 times more frequent code deployments

  2. 106 times faster lead time from commit to deploy

  3. 2,604 times faster time to recovery from incidents

  4. 7 times lower change failure rate 

SDM integrates the five pillars of modern software delivery: 1) continuous integration (CI), 2) continuous delivery (CD), 3) feature management, 4) feature management, and, 5) value stream management across multiple teams, multiple processes, and multiple tools and stacks. SDM sits atop all DevOps tools to help analyze data and signals. These signals help measure the success of the changes that are taking place with software development and deployment. By enabling everyone to see the performance metrics, they’re able to see the effect of their efforts.


During CloudBees Connect, CloudBees announced the integration of their SDM platform with Google Cloud Build and Tekton pipelines. The announcement is consistent with CloudBees commitment to help those involved in software delivery and development to become more efficient, collaborate more effectively, and share software delivery insights across the organization while continuing to use the tools that work best for the organization.


Use Case

Bob Kelly, Director of Delivery Engineering at Intercontinental Hotel Group shared his experience building a robust, repeatable DevOps practice that results in high-quality, automated certification of portfolio apps across IHG’s brands and geographies.


He noted people want change, but they don’t want to change. In order to know whether or not the changes you are making are achieving the desired result. They now include unit and functional tests in their “definition of done.”


Over the last nine months, the IHG team has accomplished:

  1. Build out the entire DevOps platform with more than 100 VMs in 9 data centers with more than 900 firewall rules

  2. Built 40 pipelines for dev teams and are working on more

  3. Created an inner sourcing model where anyone can contribute to the codebase

  4. Delivered web properties to 8 data centers worldwide in less than 2 hours

  5. Delivered to on-prem and cloud platforms

  6. Working on parameterized microservices pipeline implementation

The most progressive team can complete full production releases in less than 30 minutes versus a week with every pull request going through exhaustive certification.

Drop Me a Line, Let Me Know What You Think

© 2020 by Tom Smith | ctsmithiii@gmail.com | @ctsmithiii