Deploying, maintaining, and migrating production n8n on EKS
A team ran n8n in queue mode on Amazon EKS with separate workers and webhook processors, syncing secrets from AWS Secrets Manager via External Secrets Operator. They then moved from a community Helm chart to the official n8n Helm chart as a full cutover, keeping PostgreSQL, Redis, S3, IAM and secrets in place. The main pod's persistence was disabled because they found no state on that volume that needed to survive a restart.
What changes for platform teams
Platform teams running n8n on Kubernetes can see a concrete migration path from the community chart to the official chart, including the trade-off that main pods move to a Deployment with a shared PVC that gp3 RWO volumes cannot serve across nodes. The article's sample values repository shows queue mode, HPA and ingress configuration for external Postgres and Redis.
Read the original at CNCF Blog ↗
written by claude-haiku-5-5, prompt 7cbe154, checked by publish rules