ZeroDownTime Kubernetes platform - https://kubezero.com
Go to file
2020-09-08 15:34:16 +01:00
charts Fix fluentd parsing of json 2020-09-08 15:34:16 +01:00
deploy Add fluent-bit support to kuberzero-logging, istio fixes 2020-08-22 18:27:18 +01:00
docs Move scrape username to its own secret as eck operator cleans up otherwise 2020-09-04 01:13:39 +01:00
scripts Add fluent-bit support to kuberzero-logging, istio fixes 2020-08-22 18:27:18 +01:00
.gitattributes Initial commit 2020-05-01 14:57:31 +01:00
.gitignore Block Chart.lock as it breaks helm v3 in argo for deps 2020-05-07 00:36:00 +01:00
.helmdocsignore New cleaned up KubeZero layout 2020-06-25 13:52:07 +01:00
CHANGES.md Add custom prometehus-operator settings 2020-08-06 11:52:16 +01:00
LICENSE.md Initial commit 2020-05-01 14:57:31 +01:00
README.md Update docs 2020-08-26 11:41:28 +01:00

KubeZero - Zero Down Time Kubernetes platform

KubeZero is a pre-configured collection of components deployed onto a bare Kubernetes cluster. All chosen components are 100% organic OpenSource.

Design goals

  • Cloud provider agnostic, bare-metal / self-hosted possible
  • No vendor lock in
  • No closed source solutions
  • No premium services / subscriptions required
  • Staying to upstream projects as close as possible
  • Minimal custom code
  • Work within each community / give back

General

  • Container runtime cri-o rather than Docker for improved security and performance

Control plane

  • support for single node control plane for small clusters / test environments to reduce costs
  • access to control plane from within the VPC only by default ( VPN access required for Admin tasks )
  • controller nodes are used for various platform admin controllers / operators to reduce costs and noise on worker nodes
  • integrated ArgoCD Gitops controller

AWS IAM access control

  • Kiam allowing IAM roles per pod
  • IAM roles are assumed / requested and cached on controller nodes for improved security
  • blocking access to meta-data service on all nodes
  • IAM roles are maintained/ automated and tracked via CFN templates

Network

  • Calico using VxLAN incl. increased MTU
  • allows way more containers per worker
  • isolates container traffic from VPC by using VxLAN overlay
  • no restrictions on IP space / sizing from the underlying VPC architecture

Storage

  • flexible EBS support incl. zone awareness
  • EFS support via automated EFS provisioning for worker groups via CFN templates
  • local storage provider for latency sensitive high performance workloads

Ingress

  • AWS Network Loadbalancer and Istio Ingress controllers
  • No additional costs per exposed service
  • Automated SSL Certificate handling via cert-manager incl. renewal etc.
  • support for TCP services
  • Client source IP available to workloads via HTTP header
  • optional full service mesh

Metrics

  • Prometheus support for all components
  • automated service discovery allowing instant access to common workload metrics
  • Preconfigured community maintained Grafana dashboards for common services
  • Preconfigured community maintained Alerts

Logging

  • all container logs are enhanced with Kubernetes metadata to provide context for each message
  • flexible ElasticSearch setup via ECK operator to ease maintenance and reduce required admin knowledge, incl automated backups to S3
  • Kibana allowing easy search and dashboards for all logs, incl. pre configured index templates and index management to reduce costs
  • fluentd central log ingress service allowing additional parsing and queuing to improved reliability
  • lightweight fluent-bit agents on each node requiring minimal resources forwarding logs secure via SSL to fluentd