Senior Manager of DevOps and SecurityLogMeIn
Sebastian Daehne is a Senior Manager of DevOps and Security at LogMeIn. Sebastian has deployed software for the past nineteen years. Daehne began his career as a C++ developer and eventually became the leader of other Software Engineering teams. Today he assists teams across his organization with their DevOps journeys. Daehne leverages his experience with containers and Kubernetes to enable automation, cloud usage, and deployments at scale.
Hero’s Current Activity
The Journey to Mastering Continuous Delivery:
- Migrated to a container-native architecture using mainly Shell Scripting to orchestrate deployments.
- Long-running and complex deployment scenarios.
- Enterprise support for the Continuous Delivery process.
- Enhance their deployments and make them more reliable.
- Leverage Kubernetes and other tools to deploy and configure infrastructure at scale.
- The key to continuous delivery for LMI was to reduce the risk of complex deployments. “At the time Jenkins required a high level of engineering effort” to conform to deployments that required a high cardinality of installations(gradually roll out services). They needed a tool that would allow them to model and configure their deployment steps.
- Find a dedicated CD platform that manages deployment business logic that they would otherwise need to implement in Jenkins.
- Scale through self-service deployments: “We wanted to make it easier for the developer to just hit a button and see that it deploys”
- Have a user interface experience that makes change approvals easier to understand and manage.
- Unopinated pipelines and flexibility: “Jenkins is not really used for very long-running jobs. In Harness, I can have a deployment that takes a week if I need to, and there’s no problem.”
Future of DevOps
- DevOps’s purpose is to enable engineering teams to develop and deploy their own applications. In order to do this organizations should pursue abstraction and automation.
- We saw this happen in the operations community as people found alternatives to shell scripts. “There’s nothing wrong with a good set of shell scripts, you’re not getting slower by that. But your competition is getting faster.” Daehne recommends investing in tools that work for your use cases and will help you accelerate. Leaders will trade engineering efforts for tools that provide a service.
- Kubernetes can help provide the needed abstraction. Nothing is more frustrating than having a pipeline fail because of some weird backend issue, and having to start over. Kubernetes allows you to have independent deployments that can restart automatically with limited impact on hardware failure. In my opinion, the industry will adopt Kubernetes on and off-prem because it allows more flexibility in the deployment process.
- Keep in mind to adjust roles, responsibilities and required skill sets along the way – the cloud is not just a data center somewhere else – it offers a lot of interfaces for automation.
“In Harness, I can have a deployment that takes a week if I need to, and there’s no problem”
An educator turned software development engineer, David Sudia is a Senior DevOps Engineer at GoSpotCheck…
Glenn Turner Jr.
Glenn Turner Jr. is a Senior Software Engineer at Discover Financial Services. He delivers software and creates best practices for continuous delivery.
About the Heroes Program
To celebrate the leaders, the visionaries, and the contributors, The Heroes of CD program shares the accomplishments of industry professionals…