rancher-partner-charts/charts/trilio/k8s-triliovault-operator

K8s-TrilioVault-Operator

This operator is to manage the lifecycle of TrilioVault Backup/Recovery solution. This operator install, updates and manage the TrilioVault application.

Introduction

Prerequisites

• Kubernetes 1.19+
• PV provisioner support
• CSI driver should be installed

One Click Installation

In one click install for upstream operator, a cluster scope TVM custom resource triliovault-manager is created.

helm repo add trilio-vault-operator https://charts.k8strilio.net/trilio-stable/k8s-triliovault-operator
helm install tvm trilio-vault-operator/k8s-triliovault-operator

One click install with preflight Configuration

The following table lists the configuration parameter of the upstream operator one click install feature as well as preflight check flags, their default values and usage.

Parameter	Description	Default	Example
installTVK.enabled	1 click install feature is enabled	true
installTVK.applicationScope	scope of TVK application created	Cluster
installTVK.tvkInstanceName	tvk instance name	""	"tvk-instance"
installTVK.ingressConfig.host	host of the ingress resource created	""
installTVK.ingressConfig.tlsSecretName	tls secret name which contains ingress certs	""
installTVK.ingressConfig.annotations	annotations to be added on ingress resource	""
installTVK.ingressConfig.ingressClass	ingress class name for the ingress resource	""
installTVK.ComponentConfiguration.ingressController.enabled	TVK ingress controller should be deployed	true
installTVK.ComponentConfiguration.ingressController.service.type	TVK ingress controller service type	"NodePort"
preflight.enabled	enables preflight check for tvk	false
preflight.storageClass	Name of storage class to use for preflight checks (Required)	""
preflight.cleanupOnFailure	Cleanup the resources on cluster if preflight checks fail (Optional)	false
preflight.imagePullSecret	Name of the secret for authentication while pulling the images from the local registry (Optional)	""
preflight.limits	Pod memory and cpu resource limits for DNS and volume snapshot preflight check (Optional)	""	"cpu=600m,memory=256Mi"
preflight.localRegistry	Name of the local registry from where the images will be pulled (Optional)	""
preflight.nodeSelector	Node selector labels for pods to schedule on a specific nodes of cluster (Optional)	""	"key=value"
preflight.pvcStorageRequest	PVC storage request for volume snapshot preflight check (Optional)	""	"2Gi"
preflight.requests	Pod memory and cpu resource requests for DNS and volume snapshot preflight check (Optional)	""	"cpu=300m,memory=128Mi"
preflight.volumeSnapshotClass	Name of volume snapshot class to use for preflight checks (Optional)	""
preflight.logLevel	Log Level for the preflight run (Default: "INFO")	""
preflight.imageTag	Image tag to use for the preflight image (Default: latest)	""

Check the TVM CR configuration by running following command:

kubectl get triliovaultmanagers.triliovault.trilio.io triliovault-manager -o yaml

Once the operator pod is in running state, the TVK pods getting spawned. Confirm the TVK pods are up.

Note:

If preflight check is enabled and helm install fails, check pre-install helm hook pod logs for any failure in preflight check. Do the following steps:

First, run this command:

kubectl get pods -n <helm-release-namespace>

The pod name should start with <helm-release-name>-preflight-job-preinstall-hook. Check the logs of the pod by the following command:

kubectl logs -f <pod-name> -n <helm-release-namespace>

The failed preflight job is not cleaned up automatically right after failure. If the user cluster version is 1.21 and above, the job will be cleaned up after 1 hour so user should collect any failure logs within 1 hr of job failure. For cluster version below 1.21, user has to clean up failed preflight job manually.

To delete the job manually, run the following command:

kubectl delete job -f <job-name> -n <helm-release-namespace>

where job name should also start with <helm-release-name>-preflight-job-preinstall-hook

Also, due to a bug at helm side where auto deletion of resources upon failure doesn't work, user needs to clean the following resources left behind to be able to run preflight again, until the bug is fixed from their side, after which this step will be handled automatically. Run the following command to clean up the temporary resources:

1. Cleanup Service Account:

   kubectl delete sa <helm-release-name>-preflight-service-account -n <helm-release-namespace>
   

2. Cleanup Cluster Role Binding:

   kubectl delete clusterrolebinding <helm-release-name>-<helm-release-namespace>-preflight-rolebinding
   

3. Cleanup Cluster Role:

   kubectl delete clusterrole <helm-release-name>-<helm-release-namespace>-preflight-role
   

Manual Installation

To install the operator on local setup just run the latest helm charts inside this repo

helm repo add trilio-vault-operator https://charts.k8strilio.net/trilio-stable/k8s-triliovault-operator
helm install tvm trilio-vault-operator/k8s-triliovault-operator

Now, create a TrilioVaultManager CR to install the TrilioVault for Kubernetes. You can provide the custom configurations for the TVK resources as follows:

apiVersion: triliovault.trilio.io/v1
kind: TrilioVaultManager
metadata:
  labels:
    triliovault: k8s
  name: tvk
spec:
  trilioVaultAppVersion: latest
  applicationScope: Cluster
  # User can configure tvk instance name
  tvkInstanceName: tvk-instance
  # User can configure the ingress hosts, annotations and TLS secret through the ingressConfig section
  ingressConfig:
    host: "trilio.co.in"
    tlsSecretName: "secret-name"
  # TVK components configuration, currently supports control-plane, web, exporter, web-backend, ingress-controller, admission-webhook.
  # User can configure resources for all componentes and can configure service type and host for the ingress-controller
  componentConfiguration:
    web-backend:
      resources:
        requests:
          memory: "400Mi"
          cpu: "200m"
        limits:
          memory: "2584Mi"
          cpu: "1000m"
    ingress-controller:
      enabled: true
      service:
        type: LoadBalancer

Apply the Custom Resource

Apply TVM.yaml:

kubectl create -f TVM.yaml

Check TVK Install

Check that the pods were created:

kubectl get pods

NAME                                                        READY   STATUS    RESTARTS   AGE
k8s-triliovault-admission-webhook-6ff5f98c8-qwmfc           1/1     Running   0          81s
k8s-triliovault-backend-6f66b6b8d5-gxtmz                    1/1     Running   0          81s
k8s-triliovault-control-plane-6c464c5d78-ftk6g              1/1     Running   0          81s
k8s-triliovault-exporter-59566f97dd-gs4xc                   1/1     Running   0          81s
k8s-triliovault-ingress-nginx-controller-867c764cd5-qhpx6   1/1     Running   0          18s
k8s-triliovault-web-967c8475-m7pc6                          1/1     Running   0          81s
tvm-k8s-triliovault-operator-66bd7d86d5-dvhzb               1/1     Running   0          6m48s

Check that ingress controller service is of type LoadBalancer:

NAME                                                 TYPE           CLUSTER-IP     EXTERNAL-IP      PORT(S)                      AGE
k8s-triliovault-admission-webhook                    ClusterIP      10.7.243.24    <none>           443/TCP                      129m
k8s-triliovault-ingress-nginx-controller             LoadBalancer   10.7.246.193   35.203.155.148   80:30362/TCP,443:32327/TCP   129m
k8s-triliovault-ingress-nginx-controller-admission   ClusterIP      10.7.250.31    <none>           443/TCP                      129m
k8s-triliovault-web                                  ClusterIP      10.7.254.41    <none>           80/TCP                       129m
k8s-triliovault-web-backend                          ClusterIP      10.7.252.146   <none>           80/TCP                       129m
tvm-k8s-triliovault-operator-webhook-service         ClusterIP      10.7.248.163   <none>           443/TCP                      130m                  123m

Check that ingress resources has the host defined by the user:

NAME              CLASS                           HOSTS   ADDRESS          PORTS   AGE
k8s-triliovault   k8s-triliovault-default-nginx   *       35.203.155.148   80      129m

You can access the TVK UI by hitting this address in your browser: https://35.203.155.148

Delete

kubectl delete -f TVM.yaml

Uninstall

To uninstall/delete the operator helm chart :

helm uninstall tvm

TrilioVaultManager compatibility

We maintain the version parity between the TrilioVaultManager(upstream operator) and TrilioVault for Kubernetes. Whenever user wants to upgrade to the new version, should use the same version for upstream operator and Triliovault for Kubernetes.