For the last three years we have been taking part in a large-scale IT project at a corporate enterprise in Germany that was meant to be the first venture tackling all the latest buzz in the IT world including most prominently DevOps and continuous deployment in order to achieve a minimal time-to-market. Among the others are Agile, microservices, “the cloud”, and more. During this time we saw the project grow from 50 people in 6 teams to almost 500 in 40 teams and back to 150. We observed this huge enterprise company (and its employees) struggle to adjust its age-old ways and mindset and almost aborting the project in the process, twice.
And most of all, we failed. Hard. By all measurable goals set for this project, we failed to adopt a DevOps culture, we failed to establish independent teams, we failed to maintain a high release cadence and we almost failed at delivering a working product at all. However, we learned valuable lessons about organizations, culture, processes and adopting DevOps in large enterprises. In fact and perhaps surprising to some, almost none of the major issues and mistakes we made on the way are tech-related. Many challenges arise from the unique circumstance that the company is not at all a tech-savvy start-up (but rather a very large enterprise with hundreds of thousands of employees), and IT is not (yet) part of its core business unlike say Amazon or Microsoft. Therefore, this talk is as much an observation of the great lengths and weird processes to which large organizations will go as it is a failure story of adopting DevOps and continuous deployment in particular.
In this talk I will briefly walk you through the wild ride we have been through, point out the key lessons we learned and hopefully encourage you to do better in your own projects.