i s i s n o t a b o u t provisioning a VM, setting up a Kubernetes cluster or replacing a faulty HDD … … this would be on the infrastructure team. appops
DEVELOP IN A DISTRIBUTED SETUP 9 hackernoon.com/how-we-develop-in-and-with-distributed-systems-66c71e918735 remote local cluster dev cluster dev cluster dev dev cluster pure off-line proxied live pure on-line
DEVELOP IN A DISTRIBUTED SETUP 10 hackernoon.com/how-we-develop-in-and-with-distributed-systems-66c71e918735 DC/OS Vagrant Docker Mac/Windows Minikube, Minishift no need to maintain cluster no costs off-line and fast dev cycles limited scale not really a distributed system not all functionality might be available pure off-line
DEVELOP IN A DISTRIBUTED SETUP 11 hackernoon.com/how-we-develop-in-and-with-distributed-systems-66c71e918735 DC/OS Tunnel kubectl proxy/apiserver proxy almost the real thing (distributed system) fast dev cycles have to maintain cluster/cost limited off-line development proxied
DEVELOP IN A DISTRIBUTED SETUP 12 hackernoon.com/how-we-develop-in-and-with-distributed-systems-66c71e918735 Artifactory + Bamboo + Marathon GitLab CR + Jenkins + OpenShift Quay.io + TeamCity + Tectonic the real thing (distributed system) have to maintain cluster/cost very limited off-line development slow dev cycles live
DEVELOP IN A DISTRIBUTED SETUP 13 hackernoon.com/how-we-develop-in-and-with-distributed-systems-66c71e918735 Google Cloud Shell/GCR/GKE Katacoda Koding the real thing (distributed system) low/no costs scales well no off-line development limited customization/control pure on-line