Skip to content


beryl  cloudflare  cloudflare pages  wrangler  container  docker  registry  harbor  Dockerfile  kaniko  bytebase  gitops  mongodb  pymongo  mdbc  kubernetes  operator  os  debian  preseed  automation  git  github  credentials  gh  git tag  tips  gitlab  cicd  gitlab runner  glab  docker-compose  graphql  mutation  ollama  letta  memory  hlv3  ha  etcd  calico  flux  sops  gateway  nginx  nginx-gateway-fabric  certmanager  gitlab-runner  minio  s3  monitoring  kube-prometheus  prometheus  alertmanager  grafana  node-exporter  memcached  loki  promtail  keycloak  postgres  testbed  homelab  cilium  proxmox  cloud-init  fluxcd  cert-manager  encryption  longhorn  Kubernetes  index  dns  coredns  k9s  krew  storage  postgres-operator  hyper-v  helm  metallb  kubernetes gateway  ngf  directpv  logging  lab  reverse proxy  mfa  llm  open-webui  mcp  markdown  wip  mise  python  venv  go  node  npm  snmp  multicast  nvim  lazyvim  obsidian  ansible  poetry  regex  freenginx  mercurial  tls  tls passthrough  tls offload  bash  zsh  linux  ssg  hugo  mkdocs  tmux  tpm  pomodoro  vagrant  windows  virtualization  powershell  winget  windows terminal  windows 11 pro  oh-my-posh  nerdfonts  scoop 


longhorn

#kubernetes   #storage   #longhorn  

Table of Content

longhorn

https://longhorn.io/

Cloud native distributed block storage for Kubernetes

testing on hyper-v kubernetes cluster

  • create x4 debian VMs
    • Debian 12.4 netinst, ssh headless
    • 2G memory, dynamic
    • x4 vcpu
    • 32BG disk
    • lvm, guided
    • hostnames host1, host2, host3, and host4
  • install things required for kubernetes
  • install packages to use longhorn
    • apt install open-iscsi -y
    • systemctl enable iscsid --now
  • kubeadm init to form a cluster
  • install calico as network addon
  • install metallb
  • install nginx gateway fabric
  • install longhorn
  • add gateway and httproute in longhorn-system namespace for longhorn-frontend svc to access longhorn UI
    • use ldns to create appropriate fqdn for the longhorn UI
  • create x3 32GB disk and attach to VMs host2, host3, and host4
  • prep and mount additional disk
    • fdisk -l to confirm /dev/sd?
    • assuming it's /dev/sdb...
      • initialize by deleting the existing partitions if any using fdisk (sudo fdisk /dev/sdb, d to delete partitions, and w to write and exit)
    • pvcreate /dev/sdb
      • try sudo wipefs --all /dev/sdb if pvcreate fails
    • vgcreate vg2 /dev/sdb
    • lvcreate -n disk2 -l 100%FREE vg2
    • mkfs.ext4 /dev/vg2/disk2
    • check the mapper ls -1 /dev/mapper
    • edit /etc/fstab
      • /dev/mapper/{mamper for the new lvm} /mnt/disk2 ext4 errors=remount-ro 0 1
      • mkdir -p /mnt/disk2
      • systemctl daemon-reload
      • mount -a
  • access Longhorn UI
    • navigate to "node" tab
      • find "edit node and disks" under the operation column in the table for hosts 2, 3, and 4
      • set "disabled" to not schedule for the default disk
      • add disk, specifying the reservation size 0Gi and mount location /mnt/disk2, and set to enabled for scheduling
      • as for host1,
    • navigate to "volume" tab
      • create a new volume v1 with 8Gi
      • attach to host2
      • create PV/PVC from the operation menu of the volume
      • kubectl get pv and kubectl get pvc -A to confirm

Test further to add a pod on different hosts that use the PVC created.

removing disk

https://longhorn.io/docs/1.8.0/maintenance/maintenance/#removing-a-disk

  • on frontend-ui
    • navigate to the node menu and open edit node and disks menu
    • disable scheduling on the disk to remove
    • evict all replicas on the disk by setting true on eviction requested and save and wait
      • confirm that the replicas count goes down to 0 for this disk

removing node

https://longhorn.io/docs/1.8.0/maintenance/maintenance/#removing-a-node

  • on frontend-ui
    • navigate to the node menu and open edit node and disks menu
    • set node scheduling to disabled and eviction requested to true and save and wait
      • confirm that the replicas count goes down to 0 for this node
      • all the workloads should be migrated to other nodes
    • select the node and delete from the menu
      • the delete button is grayed out with hint text saying the node must be deleted from the kubernetes cluster first
      • if you delete the node kubectl delete node {node_to_delete}, it automatically goes away from the longhorn UI as well