Production Practice For Large-Scale Financial Application Platform In China Merchants Bank
Talkat KubeCon + CloudNativeCon North America 2022 · on Thursday, 2022-10-27· at 11:00· Virtual Platform Only·
summary
China Merchants Bank runs a large private cloud with VMs, Kubernetes, serverless and Hadoop and builds an application uniform control plane using KubeVela to decouple apps from runtimes. They define component types (frontend, webservice, VM, task, API gateway), implement canary traffic 0→100% with traffic-tagging and three-step validation plus rollback, wire OpenTelemetry → Prometheus → Grafana for SLIs/SLOs, and publish add-ons (shown by deploying Java/Tomcat and a ClickHouse operator). Demos succeed and the platform reports reduced delivery complexity and consistent rollout/observability.
AI-written summary · about
chapters (18)
- 0:00Introduction
- 1:05Talk overview
- 1:27China Merchants Bank cloud journey
- 3:36Financial platform challenges
- 4:41Application model with OAM
- 6:27Canary release workflow
- 9:23Observability with OpenTelemetry
- 11:32Cloud native platform design
- 13:20KubeVela features
- 15:34KubeVela setup and install
- 16:58KubeVela UI walkthrough
- 17:40Observability add-ons in KubeVela
- 19:27Monitoring apps and resources
- 20:54Custom component definitions
- 23:34ClickHouse add-on build
- 26:29ClickHouse deployment and dashboard
- 28:41KubeVela roadmap
- 29:46Closing remarks
related talks
42mScaling Digital Employees at China Merchants Bank in Production: A Controlled AI Agents A...
24mBanking Observability at Scale
32mLesson Learnt Running an Indian Bank on Kubernetes
56mGetting all the 99.99(9) you always wanted
35mBuilding a Bank with Kubernetes
37mCITIC Bank's Containerized Exploration Road
31mReal-World DevOps Experiences from a 165-year Old Bank
57mBuilding a bank from scratch in the Cloud