The Good, the Bad & the Ugly of Migrating Hundreds of Legacy Applications to Kubernetes
summary
QAware migrates 400 JEE monoliths onto Kubernetes/OpenShift on AWS, using questionnaires, static analysis (IBM migration tool, SonarQube), a migration database and Tableau to prioritize work. The target runs JVM 8, TLS 1.2 two‑way everywhere, an edge service that issues OpenID Connect identity tokens, and container patterns (sidecar, ambassador, adapter). Session sync proves hardest and is solved with Apache Ignite after rejecting sticky sessions, DB session persistence, Redis and Hazelcast. About 86–100 systems are live now and 200 target live by Q1 2018.
AI-written summary · about
chapters (14)
- 0:00Intro
- 0:53Cloud native applications
- 2:05A brave decision
- 3:59What we have learned
- 5:16The Good
- 10:12Emergent Design
- 10:44Migrating 400 Applications
- 12:21Source Architecture
- 17:24Edge Service
- 19:42Container Patterns
- 22:03Kubernetes Restrictions
- 23:21State State Files
- 31:53Security
- 33:47Ugly
related talks
30mMigrating a monolith to Cloud-Native and the stumbling blocks
44mMigrating a Monolith to Kubernetes
33mNavigating the Seas of Transformation: Migrating from EC2 to Kubernetes
20mKeynote: The Challenges of Migrating 150+ Microservices to Kubernetes
31mMigrating Legacy Infrastructure to Kubernetes: Olark’s Lessons From the Trenches
46mThe Journey of Migrating 2500 Apps to Multi-Cluster Kubernetes and Zero Trust Istio
20mApplication Modernization: From Concept to Kubernetes
35mPerforming Infrastructure Migrations at Scale