Cloud Native DevOps at the Scale of Salesforce

Session Description

Let's talk about how we migrated two masters and 60 agents, 1500 developers and 100 scrum teams from on-premise Jenkins Enterprise to CloudBees CI using AWS EKS.
Let's talk about how we Configured the Entire EKS Cluster using Terraform and we have configured CloudBees entirely using Jenkins Configuration as Code.
Let's talk about how we customized our Operation Center image, Managed Master images, and our Build Agent images.
Let's talk about how we automated our Agent Configuration and Environment Buildout using Kaniko and Helm.
Let's talk about how we put into place RBAC and integration CloudBees configuration as code with our Corporate SSO.
Let's talk about how we reduced Jenkins upgrades from 3 person 3 weeks to 1 person 1 day. which is a saving of time and resources of 500%.
Let's talk about how we codified our plugins where all plugins are installed and upgraded via Code.
Let's talk about what challenges we ran into with adopting EKS, CloudBees Configuration as Code, Docker, Kaniko and Helm.

For the legacy environment, we lacked audit trails. Now that everything is managed through code, which has allowed the Salesforce team to revert changes and make changes faster because the development environment and production environment are better integrated. In the past, feature development would involve a couple of sprints and take approximately one month. Now a feature can be released in 2 to 3 days, reducing the time for new feature deployment by 85%.