How CNET (And Friends) Use the CNCF Landscape To Run High Traffic, Dynamic, Scalable, And Cost-Effective Websites.

on YouTube ↗34m

" summary

CNET moves 45 billion edge requests a month and ~20,000 req/s at origin and runs an automated pipeline that builds sandboxed environments in about five minutes, enabling 10–20 production deploys per day. The platform wires Helm, Traefik Ingress controller, External DNS, cert-manager and Prometheus on GKE Autopilot, automates TLS issuance and DNS, uses Grafana/Alertmanager for metrics, and a cleanup tool (deckhand); the stack cuts errors and speeds validation.

AI-written summary · about

" chapters (7)

1. 0:00Start
2. 7:56Workload/Deployment
3. 8:09Networking
4. 13:27External DNS
5. 15:27Cert Manager
6. 19:21Deckhand (internal)
7. 20:13Observability

" related talks

• 24mIntro: CNCF CILucina Stricko, Denver Williams · KubeCon + CloudNativeCon Europe 2019
• 22mCreating a Micro Open-Source Community with HelmKatie Gamanji · KubeCon + CloudNativeCon North America 2019
• 28mArchitecting Kubernetes Clusters for High Traffic SitesAlexander Serebryakov · Infobip Shift 2023
• 13mKeynote: A Journey to a Centralized, Globally Distributed Platform – Katie GamanjiKatie Gamanji · KubeCon + CloudNativeCon Europe 2019
• 20mOperating High Traffic Websites on KubernetesSalman Iqbal · KCD Pakistan 2023
• 26mCNCF Cross-Cloud CI IntroTaylor Carpenter, Denver Williams · KubeCon + CloudNativeCon Europe 2018
• 17mIntro: CNCF Cross-Cloud CIW. Watson · KubeCon + CloudNativeCon China 2018
• 26mDeep Dive: CNCF CILucina Stricko, Denver Williams · KubeCon + CloudNativeCon Europe 2019