Terraform configuration for deploying Traefik ingress controller in a Kubernetes cluster
Copy terraform.tfvars.example file to terraform.tfvars and set input variables values as per your needs. Then initialize Terraform with init command:
terraform init -backend-config "bucket=$BUCKET_NAME" -backend-config "prefix=apps/traefik" -backend-config "region=$REGION"$REGIONshould be replaced with a region name.$BUCKET_NAMEshould be replaced with a GCS Terraform state storage bucket name.
To apply Terraform plan, run:
terraform applySome of the included services require persistent storage, configured through Persistent Volumes that specify which disks your cluster has access to.
Storage changes after installation need to be manually handled by your cluster administrators. Automated management of these volumes after installation is not handled by the deployment scripts.
IMPORTANT: you may experience a total data loss if these changes are not applied properly. Specifically, if you don't change the default
DeletereclaimPolicy for PersistentVolumes toRetain, the inderlying Google Persistent Disk will be completely destroyed by GCE upon destruction of a Kubernetes application stack.
Find the volumes/claims that are being used, and change the reclaimPolicy for each from Delete to Retain:
$ kubectl get pv | grep datadir-consul
pvc-d4184653-179a-11e9-ad5f-42010a80029a 1Gi RWO Delete Bound kube-system/datadir-consul-0 standard 9h
$ kubectl patch pv pvc-d4184653-179a-11e9-ad5f-42010a80029a -p "{\"spec\":{\"persistentVolumeReclaimPolicy\":\"Retain\"}}"
persistentvolume "pvc-d4184653-179a-11e9-ad5f-42010a80029a" patched
$ kubectl get pv -n kube-system | grep datadir-consul
pvc-d4184653-179a-11e9-ad5f-42010a80029a 1Gi RWO Retain Bound kube-system/datadir-consul-0 standard 9hAfter you uninstall this plan from the cluster, and you are completely sure you don't need its persisten volumes anymore, you can delete them using following commands:
$ kubectl get pvc | grep datadir-consul
datadir-consul-0 Bound pvc-d4184653-179a-11e9-ad5f-42010a80029a 1Gi RWO standard 15h
$ kubectl delete pvc datadir-consul-0 -n kube-system
persistentvolumeclaim "datadir-consul-0" deleted
$ gcloud compute disks list --filter="-users:*"
NAME ZONE SIZE_GB TYPE STATUS
gke-dev-a3f54e52-dynam-pvc-e41dcd6c-179a-11e9-ad5f-42010a80029a us-central1-a 1 pd-standard READY
$ gcloud compute disks delete gke-dev-a3f54e52-dynam-pvc-e41dcd6c-179a-11e9-ad5f-42010a80029a --zone=us-central1-a
The following disks will be deleted:
- [gke-dev-a3f54e52-dynam-pvc-e41dcd6c-179a-11e9-ad5f-42010a80029a]
in [us-central1-a]
Do you want to continue (Y/n)? y
Deleted [https://www.googleapis.com/compute/v1/projects/mtm-default-1/zones/us-central1-a/disks/gke-dev-a3f54e52-dynam-pvc-e41dcd6c-179a-11e9-ad5f-42010a80029a].
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent