velero/docs/restic.md

Restic Integration

As of v0.9.0, Ark has support for backing up and restoring Kubernetes volumes using a free open-source backup tool called restic.

Ark has always allowed you to take snapshots of persistent volumes as part of your backups, provided you’re using one of the supported cloud providers’ block storage offerings (Amazon EBS Volumes, Azure Managed Disks, Google Persistent Disks). And, since v0.6.0, we’ve had a plugin model which enables anyone to easily implement additional object and block storage backends, outside of the main Ark repository.

We integrated restic with Ark so that users have an out-of-the-box solution for backing up and restoring almost any type of Kubernetes volume*. We view this as a new capability for Ark, rather than a replacement for existing functionality. If you're running on AWS, and taking EBS snapshots as part of your regular Ark backups, great! There's no need to switch to using restic. However, if you've been waiting for a snapshot plugin to be implemented for your storage platform, or you're using EFS, AzureFile, NFS, emptyDir, local, or any other volume type that doesn't have a native snapshot concept, restic may be for you.

Additionally, since restic is not tied to a specific storage platform, this integration paves the way for future work to enable cross-volume-type data migrations. Stay tuned as this evolves!

* hostPath volumes are not supported, but the new local volume type is!

Setup

This setup guide assumes you already have a working Ark v0.8.1+ installation. If not, go here for instructions. It also assumes you have the latest release tag of the Ark repo cloned locally.

Pre-Install Steps - Upgrading from a Previous v0.9 Alpha

If you're upgrading from a previous pre-release version of Ark with restic, before you proceed you'll need to:

• manually delete all of the repositories/data from your existing restic bucket
• delete all Ark backups from your cluster using ark backup delete
• delete all secrets named ark-restic-credentials across all namespaces in your cluster

Instructions

1. Download an updated Ark client from the latest release, and move it to a location within your PATH.

2. From the Ark root directory, run the following to create new custom resource definitions:

   kubectl apply -f examples/common/00-prereqs.yaml
   

3. Run one of the following for your platform to create the daemonset:

   • AWS: kubectl apply -f examples/aws/20-restic-daemonset.yaml
   • Azure: kubectl apply -f examples/azure/20-restic-daemonset.yaml
   • GCP: kubectl apply -f examples/gcp/20-restic-daemonset.yaml
   • Minio: kubectl apply -f examples/minio/30-restic-daemonset.yaml

4. Update the image tag on the Ark daemonset and deployment to match the release version you used in Step 1 (e.g. v0.9.0-alpha.3):

   kubectl -n heptio-ark set image deployment/ark ark=gcr.io/heptio-images/ark:<RELEASE_VERSION>
   kubectl -n heptio-ark set image daemonset/restic ark=gcr.io/heptio-images/ark:<RELEASE_VERSION>
   

5. Create a new bucket for restic to store its data in, and give the heptio-ark IAM user access to it, similarly to the main Ark bucket you've already set up. Note that for now, this must be a different bucket than the main Ark bucket, but we plan to remove this limitation in a future release.

6. Uncomment resticLocation in your Ark config and set the value appropriately, then apply:

   • AWS: kubectl apply -f examples/aws/00-ark-config.yaml
   • Azure: kubectl apply -f examples/azure/10-ark-config.yaml
   • GCP: kubectl apply -f examples/gcp/00-ark-config.yaml
   • Minio: kubectl apply -f examples/minio/10-ark-config.yaml

   Note that resticLocation may either be just a bucket name, e.g. my-restic-bucket, or a bucket name plus a prefix under which you'd like the restic data to be stored, e.g. my-restic-bucket/ark-repos.

You're now ready to use Ark with restic.

Backing Up

1. Run the following for each pod containing a volume that you'd like to back up using restic:

   kubectl -n YOUR_POD_NAMESPACE annotate pod/YOUR_POD_NAME backup.ark.heptio.com/backup-volumes=YOUR_VOLUME_NAME_1,YOUR_VOLUME_NAME_2,...
   

   The volume names specified in the annotation should be the names of the volumes within the pod spec.

   For example, for the following pod:

   apiVersion: v1
   kind: Pod
   metadata:
     name: sample
     namespace: foo
   spec:
     containers:
     - image: k8s.gcr.io/test-webserver
       name: test-webserver
       volumeMounts:
       - name: pvc-volume
         mountPath: /volume-1
       - name: emptydir-volume
         mountPath: /volume-2
     volumes:
     - name: pvc-volume
       persistentVolumeClaim: 
         claimName: test-volume-claim
     - name: emptydir-volume
       emptyDir: {}
   

   You'd run:

   kubectl -n foo annotate pod/sample backup.ark.heptio.com/backup-volumes=pvc-volume,emptydir-volume
   

   This annotation can also be provided in the pod template spec if using a deployment, daemonset, etc. to manage your pods.

2. Take an Ark backup as usual:

   ark backup create NAME OPTIONS...
   

3. When the backup has completed, view information about your pod volume backups:

   kubectl -n heptio-ark get podvolumebackups -l ark.heptio.com/backup-name=YOUR_BACKUP_NAME -o yaml
   

Restoring

Restore from your Ark backup as usual:

ark restore create --from-backup BACKUP_NAME OPTIONS...

Limitations

There are several limitations that users should be aware of:

• As mentioned previously, you cannot use the main Ark bucket for storing restic backups. We plan to address this in a future release.
• hostPath volumes are not supported. Local persistent volumes are, though!
• Those of you familiar with restic may know that it encrypts all of its data. We've decided to use a static, common encryption key for all restic repositories created by Ark. This means that anyone who has access to your bucket can decrypt your restic backup data. You should ensure that you are limiting access to the restic bucket appropriately. We plan to implement full Ark backup encryption, including securing the restic encryption keys, in a future release.

Troubleshooting

If something's not working as expected, there are several places you can go to for more information on what might be happening.

Are your Ark server and daemonset pods running?

kubectl get pods -n heptio-ark

Does your restic repository exist, and is it ready?

ark restic repo get

ark restic repo get REPO_NAME -o yaml

Are there any errors in your Ark backup/restore?

ark backup describe BACKUP_NAME
ark backup logs BACKUP_NAME

ark restore describe RESTORE_NAME
ark restore logs RESTORE_NAME

What is the status of your pod volume backups/restores?

kubectl -n heptio-ark get podvolumebackups -l ark.heptio.com/backup-name=BACKUP_NAME -o yaml

kubectl -n heptio-ark get podvolumerestores -l ark.heptio.com/restore-name=RESTORE_NAME -o yaml

Is there any useful information in the Ark server or daemon pod logs?

kubectl -n heptio-ark logs deploy/ark
kubectl -n heptio-ark logs DAEMON_POD_NAME

NOTE: You can increase the verbosity of the pod logs by adding --log-level=debug as an argument to the container command in the deployment/daemonset pod template spec.

Details of Backup & Restore Process with Restic

This section describes in detail how backups/restores with Ark and restic work.

As background, we've introduced three new custom resource definitions and associated controllers:

• ResticRepository - represents/manages the lifecycle of Ark's restic repositories. Ark creates a restic repository per namespace, when the first restic backup for a namespace is requested. The controller for this custom resource executes restic repository lifecycle commands -- restic init, restic check, and restic prune.

  You can see information about your Ark restic repos with ark restic repo get.

• PodVolumeBackup - represents a restic backup of a volume in a pod. The main Ark backup process creates one or more of these when it finds an annotated pod. Each node in the cluster runs a controller for this resource (in a daemonset) that handles the PodVolumeBackups for pods on that node. The controller executes restic backup commands to backup pod volume data.

• PodVolumeRestore - represents a restic restore of a pod volume. The main Ark restore process creates one or more of these when it encounters a pod that has associated restic backups. Each node in the cluster runs a controller for this resource (in the same daemonset as above) that handles the PodVolumeRestores for pods on that node. The controller executes restic restore commands to restore pod volume data.

Backups

1. The main Ark backup process checks each pod that it's backing up for the annotation specifying a restic backup should be taken (backup.ark.heptio.com/backup-volumes)
2. When found, Ark first ensures a restic repository exists for the pod's namespace, by:
   • checking if a ResticRepository custom resource already exists
   • if not, creating a new one, and waiting for the ResticRepository controller to init/check it
3. Ark then creates a PodVolumeBackup custom resource per volume listed in the pod annotation
4. The main Ark process now waits for the PodVolumeBackup resources to complete or fail
5. Meanwhile, each PodVolumeBackup is handled by the controller on the appropriate node, which:
   • has a hostPath volume mount of /var/lib/kubelet/pods to access the pod volume data
   • finds the pod volume's subdirectory within the above volume
   • runs restic backup
   • updates the status of the custom resource to Completed or Failed
6. As each PodVolumeBackup finishes, the main Ark process captures its restic snapshot ID and adds it as an annotation to the copy of the pod JSON that's stored in the Ark backup. This will be used for restores, as seen in the next section.

Restores

1. The main Ark restore process checks each pod that it's restoring for annotations specifying a restic backup exists for a volume in the pod (snapshot.ark.heptio.com/<volume-name>)
2. When found, Ark first ensures a restic repository exists for the pod's namespace, by:
   • checking if a ResticRepository custom resource already exists
   • if not, creating a new one, and waiting for the ResticRepository controller to init/check it (note that in this case, the actual repository should already exist in object storage, so the Ark controller will simply check it for integrity)
3. Ark adds an init container to the pod, whose job is to wait for all restic restores for the pod to complete (more on this shortly)
4. Ark creates the pod, with the added init container, by submitting it to the Kubernetes API
5. Ark creates a PodVolumeRestore custom resource for each volume to be restored in the pod
6. The main Ark process now waits for each PodVolumeRestore resource to complete or fail
7. Meanwhile, each PodVolumeRestore is handled by the controller on the appropriate node, which:
   • has a hostPath volume mount of /var/lib/kubelet/pods to access the pod volume data
   • waits for the pod to be running the init container
   • finds the pod volume's subdirectory within the above volume
   • runs restic restore
   • on success, writes a file into the pod volume, in an .ark subdirectory, whose name is the UID of the Ark restore that this pod volume restore is for
   • updates the status of the custom resource to Completed or Failed
8. The init container that was added to the pod is running a process that waits until it finds a file within each restored volume, under .ark, whose name is the UID of the Ark restore being run
9. Once all such files are found, the init container's process terminates successfully and the pod moves on to running other init containers/the main containers.