Shivering-Isles GitOps Infrastructure

This repository has become the centre of Shivering-Isles Infrastructure. It homes basically all deployments of software, various custom container images, various self-maintained helm charts and more.

Usage

For SI-GitLab this would look like this:

export GITLAB_TOKEN=<project access token able to write the API and repository>
flux bootstrap gitlab \
  --hostname=git.shivering-isles.com \
  --ssh-hostname=git.shivering-isles.com:2222 \
  --ssh-key-algorithm ed25519 \
  --owner=<your user / team> \
  --repository=<your repository name> \
  --path=clusters/<your cluster name>

Ideas & ToDo's

This toolchain is still under development. Before it will be used in production there are still some things left to do:

• Buy hardware for the project.
• Provide CLI container that contains all tools.
• Automate overlay network deployment (calico)
• Use encrypted overlay network (calico+wireguard)
• Automate cluster monitoring deployment (kube-prometheus)
• Automate ingress-controller deployment (ingress-nginx)
• Automate policy enforcement (kyverno) deployment
• Encrypt root filesystems for all nodes (LUKS + clevis)
• Enforce SELinux on the deployed machines
• Automate system upgrades using Kubernetes (system-upgrade-controller)
• Automate system configuration using Kubernetes (system-upgrade-controller)
• Provide an fully encrypted (handled on host level) storage class (longhorn)
• Deploy cert-manager
• Deploy credentials for cert-manager
• Automate ingress-controller default certificate deployment
• Add encrypted deployment instructions (SOPS + fluxcd)
• Integrate Renovatebot with this repository to manage updates.
• Automate Kubernetes upgrades
• Automate ingress-controller configuration for proxy-protocol
• Migrate apps to GitOps and Kubernetes
• Deploy kubelet with proper certificates
• Centralised logging (using loki/greylog/ELK/…)
• Add secure runtime class (gVisor/kata-container/…) for exposed containers
• Move to immutable base-system
• Set Priority classes for workloads (on namespace level it's done, but helm releases still need adjustment)
• Migrate to Forgejo
  • Kubernetes
    • fluxcd source
    • fluxcd webhook
    • renovate
    • Tekton
  • Restore SI-GitLab repositories
    • infrastructure-gitops
    • infrastructure
    • hcloud-dynfw
    • blog
    • …
  • Restore helm releases and alike
• Harden egress NetworkPolicies keycloak, forecastle, matrix
• Backup essential volumes using mc and client-side encryption and Object Locks
• Migrate S3 from Minio to VersityGw
  • Mastodon
  • Hedgedoc
  • gitops-docs
  • runbooks
  • postgres-operator logical backup
  • postgres-operator WAL/Base backup
  • longhorn backup
  • blog-archive
  • gitlab-runner-cache
  • shelly-firmware
  • togehter