How to Achieve Scalable Environments using Kubernetes and VMSS on Azure Cloud
summary
eToro moves dev and QA from slow VM bottlenecks to Kubernetes managed cluster for self-service personal environments and to VM Scale Sets on Azure for large test runs and multiregion production. build details include startup scripts, Jenkins-triggered image pipelines, Puppet/Ansible config, and an aggregator for test reports. results: end-to-end test suites drop from ~26 hours on one server to ~30 minutes on a cluster, deployment times fall from ~14–20 minutes to ~5–7, and typical test runs use 70–100 parallel VMs.
AI-written summary · about
chapters (21)
- 0:00Introduction
- 1:24Our Platform
- 2:38Distribution
- 3:42Puppet
- 4:37Agenda
- 8:45How many people use Kubernetes
- 9:09Why Kubernetes
- 10:10Applicant Issue
- 11:55Kubernetes Services
- 13:21Initiate Pipeline
- 15:08Jenkins Pipeline
- 16:58Kubernetes Dashboard
- 18:33Benefits
- 19:40Challenges
- 20:25VMSS
- 20:50Jenkins
- 23:54Automation
- 25:17Production Readiness
- 28:48Multiregion Architecture
- 29:46Multiregion Deployment
- 31:53Summary
related talks
33mNavigating the Seas of Transformation: Migrating from EC2 to Kubernetes
40mDevOps and Kubernetes in VSTS on Azure and Beyond
26mIntro to Cloud Provider Azure
39mKubernetes – Infinite Space for Your Tests
37mRunning Kubernetes in a Manufacturing Line
29mFrom EC2 to Kubernetes Takeaways from a Brownfield Migration
39mLessons Learned Migrating an Existing Product to a Multi Tenant Cloud Native Environment
32mDon't Start With 500 Clusters: Patterns in Scaling Multi-Cluster Kubernetes…