OADP 1.3 release notes - OADP Application backup and restore | Backup and restore

OADP 1.3.1 release notes
OADP 1.3.0 release notes

The release notes for OpenShift API for Data Protection (OADP) 1.3 describe new features and enhancements, deprecated features, product recommendations, known issues, and resolved issues.

OADP 1.3.1 release notes

The OpenShift API for Data Protection (OADP) 1.3.1 release notes lists new features and resolved issues.

New features

OADP 1.3.0 Data Mover is now fully supported

The OADP built-in Data Mover, introduced in OADP 1.3.0 as a Technology Preview, is now fully supported for both containerized and virtual machine workloads.

Resolved issues

IBM Cloud® Object Storage is now supported as a backup storage provider

IBM Cloud® Object Storage is one of the AWS S3 compatible backup storage providers, which was unsupported previously. With this update, IBM Cloud® Object Storage is now supported as an AWS S3 compatible backup storage provider.

OADP-3788

OADP operator now correctly reports the missing region error

Previously, when you specified profile:default without specifying the region in the AWS Backup Storage Location (BSL) configuration, the OADP operator failed to report the missing region error on the Data Protection Application (DPA) custom resource (CR). This update corrects validation of DPA BSL specification for AWS. As a result, the OADP Operator reports the missing region error.

OADP-3044

Custom labels are not removed from the openshift-adp namespace

Previously, the openshift-adp-controller-manager pod would reset the labels attached to the openshift-adp namespace. This caused synchronization issues for applications requiring custom labels such as Argo CD, leading to improper functionality. With this update, this issue is fixed and custom labels are not removed from the openshift-adp namespace.

OADP-3189

OADP must-gather image collects CRDs

Previously, the OADP must-gather image did not collect the custom resource definitions (CRDs) shipped by OADP. Consequently, you could not use the omg tool to extract data in the support shell. With this fix, the must-gather image now collects CRDs shipped by OADP and can use the omg tool to extract data.

Garbage collection has the correct description for the default frequency value

Previously, the garbage-collection-frequency field had a wrong description for the default frequency value. With this update, garbage-collection-frequency has a correct value of one hour for the gc-controller reconciliation default frequency.

OADP-3486

FIPS Mode flag is available in OperatorHub

By setting the fips-compliant flag to true, the FIPS mode flag is now added to the OADP Operator listing in OperatorHub. This feature was enabled in OADP 1.3.0 but did not show up in the Red Hat Container catalog as being FIPS enabled.

OADP-3495

CSI plugin does not panic with a nil pointer when csiSnapshotTimeout is set to a short duration

Previously, when the csiSnapshotTimeout parameter was set to a short duration, the CSI plugin encountered the following error: plugin panicked: runtime error: invalid memory address or nil pointer dereference.

With this fix, the backup fails with the following error: Timed out awaiting reconciliation of volumesnapshot.

For a complete list of all issues resolved in this release, see the list of OADP 1.3.1 resolved issues in Jira.

Known issues

Backup and storage restrictions for Single-node OpenShift clusters deployed on IBM Power® and IBM Z® platforms

Review the following backup and storage related restrictions for Single-node OpenShift clusters that are deployed on IBM Power® and IBM Z® platforms:

Storage: Only NFS storage is currently compatible with single-node OpenShift clusters deployed on IBM Power® and IBM Z® platforms.
Backup: Only the backing up applications with File System Backup such as kopia and restic are supported for backup and restore operations.

OADP-3787

Cassandra application pods enter in the CrashLoopBackoff status after restoring OADP

After OADP restores, the Cassandra application pods might enter in the CrashLoopBackoff status. To work around this problem, delete the StatefulSet pods with any error or the CrashLoopBackoff state after restoring OADP. The StatefulSet controller recreates these pods and it runs normally.

OADP-3767

OADP 1.3.0 release notes

The OpenShift API for Data Protection (OADP) 1.3.0 release notes lists new features, resolved issues and bugs, and known issues.

New features

Velero built-in DataMover

OADP 1.3 includes a built-in Data Mover that you can use to move Container Storage Interface (CSI) volume snapshots to a remote object store. The built-in Data Mover allows you to restore stateful applications from the remote object store if a failure, accidental deletion, or corruption of the cluster occurs. It uses Kopia as the uploader mechanism to read the snapshot data and to write to the Unified Repository.

Velero built-in DataMover is a Technology Preview feature only. Technology Preview features are not supported with Red Hat production service level agreements (SLAs) and might not be functionally complete. Red Hat does not recommend using them in production. These features provide early access to upcoming product features, enabling customers to test functionality and provide feedback during the development process.

For more information about the support scope of Red Hat Technology Preview features, see Technology Preview Features Support Scope.

Backing up applications with File System Backup: Kopia or Restic

Velero’s File System Backup (FSB) supports two backup libraries: the Restic path and the Kopia path.

Velero allows users to select between the two paths.

For backup, specify the path during the installation through the uploader-type flag. The valid value is either restic or kopia. This field defaults to kopia if the value is not specified. The selection cannot be changed after the installation.

GCP Cloud authentication

Google Cloud Platform (GCP) authentication enables you to use short-lived Google credentials.

GCP with Workload Identity Federation enables you to use Identity and Access Management (IAM) to grant external identities IAM roles, including the ability to impersonate service accounts. This eliminates the maintenance and security risks associated with service account keys.

AWS ROSA STS authentication

You can use OpenShift API for Data Protection (OADP) with Red Hat OpenShift Service on AWS (ROSA) clusters to backup and restore application data.

ROSA provides seamless integration with a wide range of AWS compute, database, analytics, machine learning, networking, mobile, and other services to speed up the building and delivering of differentiating experiences to your customers.

You can subscribe to the service directly from your AWS account.

After the clusters are created, you can operate your clusters by using the OpenShift web console. The ROSA service also uses OpenShift APIs and command-line interface (CLI) tools.

Resolved issues

ACM applications were removed and re-created on managed clusters after restore

Applications on managed clusters were deleted and re-created upon restore activation. OpenShift API for Data Protection (OADP 1.2) backup and restore process is faster than the older versions. The OADP performance change caused this behavior when restoring ACM resources. Therefore, some resources were restored before other resources, which caused the removal of the applications from managed clusters. OADP-2686

Restic restore was partially failing due to Pod Security standard

During interoperability testing, OpenShift Container Platform 4.14 had the pod Security mode set to enforce, which caused the pod to be denied. This was caused due to the restore order. The pod was getting created before the security context constraints (SCC) resource, since the pod violated the podSecurity standard, it denied the pod. When setting the restore priority field on the Velero server, restore is successful. OADP-2688

Possible pod volume backup failure if Velero is installed in several namespaces

There was a regresssion in Pod Volume Backup (PVB) functionality when Velero was installed in several namespaces. The PVB controller was not properly limiting itself to PVBs in its own namespace. OADP-2308

OADP Velero plugins returning "received EOF, stopping recv loop" message

In OADP, Velero plugins were started as separate processes. When the Velero operation completes, either successfully or not, they exit. Therefore, if you see a received EOF, stopping recv loop messages in debug logs, it does not mean an error occurred, it means that a plugin operation has completed. OADP-2176

CVE-2023-39325 Multiple HTTP/2 enabled web servers are vulnerable to a DDoS attack (Rapid Reset Attack)

In previous releases of OADP, the HTTP/2 protocol was susceptible to a denial of service attack because request cancellation could reset multiple streams quickly. The server had to set up and tear down the streams while not hitting any server-side limit for the maximum number of active streams per connection. This resulted in a denial of service due to server resource consumption.

For more information, see CVE-2023-39325 (Rapid Reset Attack)

For a complete list of all issues resolved in this release, see the list of OADP 1.3.0 resolved issues in Jira.

Known issues

CSI plugin errors on nil pointer when csiSnapshotTimeout is set to a short duration

The CSI plugin errors on nil pointer when csiSnapshotTimeout is set to a short duration. Sometimes it succeeds to complete the snapshot within a short duration, but often it panics with the backup PartiallyFailed with the following error: plugin panicked: runtime error: invalid memory address or nil pointer dereference.

Backup is marked as PartiallyFailed when volumeSnapshotContent CR has an error

If any of the VolumeSnapshotContent CRs have an error related to removing the VolumeSnapshotBeingCreated annotation, it moves the backup to the WaitingForPluginOperationsPartiallyFailed phase. OADP-2871

Performance issues when restoring 30,000 resources for the first time

When restoring 30,000 resources for the first time, without an existing-resource-policy, it takes twice as long to restore them, than it takes during the second and third try with an existing-resource-policy set to update. OADP-3071

Post restore hooks might start running before Datadownload operation has released the related PV

Due to the asynchronous nature of the Data Mover operation, a post-hook might be attempted before the related pods persistent volumes (PVs) are released by the Data Mover persistent volume claim (PVC).

GCP-Workload Identity Federation VSL backup PartiallyFailed

VSL backup PartiallyFailed when GCP workload identity is configured on GCP.

For a complete list of all known issues in this release, see the list of OADP 1.3.0 known issues in Jira.

Upgrade notes

Always upgrade to the next minor version. Do not skip versions. To update to a later version, upgrade only one channel at a time. For example, to upgrade from OpenShift API for Data Protection (OADP) 1.1 to 1.3, upgrade first to 1.2, and then to 1.3.

Changes from OADP 1.2 to 1.3

The Velero server has been updated from version 1.11 to 1.12.

OpenShift API for Data Protection (OADP) 1.3 uses the Velero built-in Data Mover instead of the VolumeSnapshotMover (VSM) or the Volsync Data Mover.

This changes the following:

The spec.features.dataMover field and the VSM plugin are not compatible with OADP 1.3, and you must remove the configuration from the DataProtectionApplication (DPA) configuration.
The Volsync Operator is no longer required for Data Mover functionality, and you can remove it.
The custom resource definitions volumesnapshotbackups.datamover.oadp.openshift.io and volumesnapshotrestores.datamover.oadp.openshift.io are no longer required, and you can remove them.
The secrets used for the OADP-1.2 Data Mover are no longer required, and you can remove them.

OADP 1.3 supports Kopia, which is an alternative file system backup tool to Restic.

To employ Kopia, use the new spec.configuration.nodeAgent field as shown in the following example:

Example

spec:
  configuration:
    nodeAgent:
      enable: true
      uploaderType: kopia
# ...

The spec.configuration.restic field is deprecated in OADP 1.3 and will be removed in a future version of OADP. To avoid seeing deprecation warnings, remove the restic key and its values, and use the following new syntax:
Example
```
spec:
  configuration:
    nodeAgent:
      enable: true
      uploaderType: restic
# ...
```

In a future OADP release, it is planned that the kopia tool will become the default uploaderType value.

Upgrading steps

Upgrading from OADP 1.2 Technology Preview Data Mover

OpenShift API for Data Protection (OADP) 1.2 Data Mover backups cannot be restored with OADP 1.3. To prevent a gap in the data protection of your applications, complete the following steps before upgrading to OADP 1.3:

Procedure

If your cluster backups are sufficient and Container Storage Interface (CSI) storage is available, back up the applications with a CSI backup.
If you require off cluster backups:
1. Back up the applications with a file system backup that uses the --default-volumes-to-fs-backup=true or backup.spec.defaultVolumesToFsBackup options.
2. Back up the applications with your object storage plugins, for example, velero-plugin-for-aws.

The default timeout value for the Restic file system backup is one hour. In OADP 1.3.1 and later, the default timeout value for Restic and Kopia is four hours.

To restore OADP 1.2 Data Mover backup, you must uninstall OADP, and install and configure OADP 1.2.

Backing up the DPA configuration

You must back up your current DataProtectionApplication (DPA) configuration.

Procedure

Save your current DPA configuration by running the following command:
Example
```
$ oc get dpa -n openshift-adp -o yaml > dpa.orig.backup
```

Upgrading the OADP Operator

Use the following sequence when upgrading the OpenShift API for Data Protection (OADP) Operator.

Procedure

Change your subscription channel for the OADP Operator from stable-1.2 to stable-1.3.
Allow time for the Operator and containers to update and restart.

Additional resources

Updating installed Operators

Converting DPA to the new version

If you need to move backups off cluster with the Data Mover, reconfigure the DataProtectionApplication (DPA) manifest as follows.

Procedure

Click Operators → Installed Operators and select the OADP Operator.
In the Provided APIs section, click View more.
Click Create instance in the DataProtectionApplication box.

Click YAML View to display the current DPA parameters.

Example current DPA

spec:
  configuration:
    features:
      dataMover:
      enable: true
      credentialName: dm-credentials
    velero:
      defaultPlugins:
      - vsm
      - csi
      - openshift
# ...

Update the DPA parameters:

Remove the features.dataMover key and values from the DPA.
Remove the VolumeSnapshotMover (VSM) plugin.

Add the nodeAgent key and values.

Example updated DPA

spec:
  configuration:
    nodeAgent:
      enable: true
      uploaderType: kopia
    velero:
      defaultPlugins:
      - csi
      - openshift
# ...

Wait for the DPA to reconcile successfully.

Verifying the upgrade

Use the following procedure to verify the upgrade.

Procedure

Verify the installation by viewing the OpenShift API for Data Protection (OADP) resources by running the following command:

$ oc get all -n openshift-adp

Example output

NAME                                                     READY   STATUS    RESTARTS   AGE
pod/oadp-operator-controller-manager-67d9494d47-6l8z8    2/2     Running   0          2m8s
pod/node-agent-9cq4q                                     1/1     Running   0          94s
pod/node-agent-m4lts                                     1/1     Running   0          94s
pod/node-agent-pv4kr                                     1/1     Running   0          95s
pod/velero-588db7f655-n842v                              1/1     Running   0          95s

NAME                                                       TYPE        CLUSTER-IP       EXTERNAL-IP   PORT(S)    AGE
service/oadp-operator-controller-manager-metrics-service   ClusterIP   172.30.70.140    <none>        8443/TCP   2m8s
service/openshift-adp-velero-metrics-svc                   ClusterIP   172.30.10.0      <none>        8085/TCP   8h

NAME                        DESIRED   CURRENT   READY   UP-TO-DATE   AVAILABLE   NODE SELECTOR   AGE
daemonset.apps/node-agent    3         3         3       3            3           <none>          96s

NAME                                                READY   UP-TO-DATE   AVAILABLE   AGE
deployment.apps/oadp-operator-controller-manager    1/1     1            1           2m9s
deployment.apps/velero                              1/1     1            1           96s

NAME                                                           DESIRED   CURRENT   READY   AGE
replicaset.apps/oadp-operator-controller-manager-67d9494d47    1         1         1       2m9s
replicaset.apps/velero-588db7f655                              1         1         1       96s

Verify that the DataProtectionApplication (DPA) is reconciled by running the following command:

$ oc get dpa dpa-sample -n openshift-adp -o jsonpath='{.status}'

Example output

{"conditions":[{"lastTransitionTime":"2023-10-27T01:23:57Z","message":"Reconcile complete","reason":"Complete","status":"True","type":"Reconciled"}]}

Verify the type is set to Reconciled.

Verify the backup storage location and confirm that the PHASE is Available by running the following command:

$ oc get backupStorageLocation -n openshift-adp

Example output

NAME           PHASE       LAST VALIDATED   AGE     DEFAULT
dpa-sample-1   Available   1s               3d16h   true

In OADP 1.3 you can start data movement off cluster per backup versus creating a DataProtectionApplication (DPA) configuration.

Example

$ velero backup create example-backup --include-namespaces mysql-persistent --snapshot-move-data=true

Example

apiVersion: velero.io/v1
kind: Backup
metadata:
  name: example-backup
  namespace: openshift-adp
spec:
  snapshotMoveData: true
  includedNamespaces:
  - mysql-persistent
  storageLocation: dpa-sample-1
  ttl: 720h0m0s
# ...