Release notes
2.1.1
January 19, 2022
In this release, PX-Backup includes enchancements in addition to the AWS Marketplace support and other backup, restore features available in the 2.1.0 release.
Improvement
Pure Storage has the following upgraded or enhanced functionality:
Improvement Number | Improvement Description |
---|---|
PB-2118 | PX-Backup now supports cross-region backup in the native GKE driver, by default. |
PB-2147 | On the AWS and GKE cloud providers, PX-Backup supports provisioning of all related PVCs consumed by the pod in the same zone, enabling the pod to come up after the restore. |
PB-2148 | PX-Backup now handles backup and restore of PVCs that use WaitForFirstConsumer binding mode in their respective storage classes. |
PB-2156 | Added support to backup and restore for OCP rdb and cephfs provisioners. |
Known Issues (Errata)
Pure Storage is aware of the following issues, check future release notes for fixes on these issues:
Issue Number | Issue Description |
---|---|
PB-2045 | PX-Backup deployment fails in the cloud while using a file based storage class.Workaround: The internal database used for PX-Backup requires block based storage. Use block based storage class for deploying PX-Backup. |
PB-2118 | When you restore a backup created in a different region, PX-Backup displays the restore operation is successful, but the pods do not run.Workaround: Cloud snapshots are region specific. To perform cross-region restore, add the BACKUP_TYPE=”Generic” parameter in the kdmp-config configmap on application clusters. |
PB-2124 | Cannot resize a PVC volume on a restored application.Workaround: To resize a PVC volume, set the AllowVolumeExpansion parameter in the destination storage class. |
PB-2129 | In the Azure cloud, if you restore a namespace with CouchDb to an alternate location, then PX-Backup restores are partial complete as the restore also includes the azure-storage-account-xx secret as part of the PVC restore This issue occurs if you do not select the Replace policy .Workaround: Select Replace policy during restore or deselect the azure-storage-account-xx secret when you backup or restore. |
PB-2142 | Backup fails on S3 backup location after installing the OpenShift Cluster Platform (OCP) on AWS with Portworx setup.Workaround: When you deploy Stork, add hostNetwork: true . |
Fixes
Pure Storage has fixed the following issues:
Issue Number | Issue Description |
---|---|
PB-1017 | Issue: RoleBindings which had Subjects with no namespaces in it were not restored.User Impact: An application using such a RoleBinding would fail to start after restoring on the destination clusterResolution: Stork will retain the Subject in a RoleBinding , if it does not have a namespace in it. |
PB-1021 | Issue: RoleBindings which had a system prefix in OCP environments were not backed up by Stork.User Impact: An application using a RoleBinding with an associated system SCC in Openshift would not start up after restore, since the associated RoleBinding was not backed up on the source side.Resolution: RoleBindings which have a prefix system:openshift:scc will be collected and backed up by Stork. |
PB-1903 | Issue: The metrics API was missing in the swagger-ui API listing.User Impact: Unable to trigger metric related API from the swagger-ui .Resolution: The metrics API is now listed in the swagger-ui . |
PB-2040 | Issue: The metrics API summary line was not present in the swagger-ui .User Impact: Documentation for the metrics API was missing in the swagger-ui .Resolution: Added the API summary for the metrics API. |
PB-2050 | Issue: PX-Backup pod logs were flooded with the following message constructing many client instances from the same exec auth config can cause performance problems during cert rotation , exhausting available network connections; clients constructed calling “aws-iam-authenticator” continuously on AWS installations.User Impact: PX-Backup pod logs filled up with log messages.Resolution: Moved this log message to tracef as part of the client-go module. |
PB-2083 | Issue: The backup failure timeout message was confusing, since it also includes the Stork installation failure information.User Impact: The error message was misleading with Stork installation failure information.Resolution: Removed the Stork installation failure message from the timeout error message. |
PB-2158 | Issue: PayG billing information is not being captured in Zuora production server due to missing clusterID and Cluster UID fields.User Impact: PX-Backup license usage page fails to load.Resolution: Fixed this issue by populating the Cluster ID with subscription ID and Cluster name as Org name . |
2.1.0
November 25, 2021
This release is specific for deploying PX-Backup on AWS Marketplace. To deploy PX-Backup outside of AWS Marketplace, use the 2.1.1 release.
New features
PX-Backup includes the following new features and enhancements:
- Offload cloud provider snapshots to backup location: You can now offload your Cloud (for example, EBS) snapshots, orchestrated by PX-Backup, to any S3 compliant storage or another region in public cloud. This enables you to maintain another recovery point in addition to the snapshots created by public cloud providers.
- Backup persistent data on files shares: Application owners can now build Kubernetes applications on scale-out file systems and apply protection policies. This supports file services from FlashBlade, Cloud based file services (for example, AWS EFS) or any NAS device that run the containarized applications.
- Backup or recover data using user roles from PX-Secure: You can now backup or recover data using the native user roles configured in the PX-Secure feature.
Known Issues (Errata)
Portworx is aware of the following issues, check future release notes for fixes on these issues:
Issue Number | Issue Description |
---|---|
OPERATOR-516 | Install Portworx using the operator, and install PX-Backup. When you backup a cluster with the Stork ServiceAccount not present, the backup fails displaying the following error:Error running PreExecRule: error executing PreExecRule for namespace csi-mysql: pods "pod-cmd-executor-09ca32f1-efe8-4b6c-89aa-0c00d741c1db" is forbidden: error looking up service account kube-system/stork-account: serviceaccount "stork-account" not found .Workaround: Create the missing ServiceAccount :apiVersion: v1 kind: ServiceAccount metadata: name: stork-account namespace: kube-system --- kind: ClusterRole apiVersion: rbac.authorization.k8s.io/v1 metadata: name: stork-role rules: - apiGroups: ["*"] resources: ["*"] verbs: ["*"] --- kind: ClusterRoleBinding apiVersion: rbac.authorization.k8s.io/v1 metadata: name: stork-role-binding subjects: - kind: ServiceAccount name: stork-account namespace: kube-system roleRef: kind: ClusterRole name: stork-role apiGroup: rbac.authorization.k8s.io |
PB-1976 | You cannot add the TKGi cluster to PX-Backup. This issue occurs because PX-Backup cannot resolve the server: https://system-test:8443 parameter in the kubeconfig file.Workaround: To add the TKGi cluster to PX-Backup, change the server parameter value in the kubeconfig file to https://10.100.200.1:443 . |
2.0.1
August 20, 2021
Improvement
Pure Storage has the following upgraded or enhanced functionality:
Improvement Number | Improvement Description |
---|---|
OC-933 | Keycloak is upgraded from version 9.0.2 to 14.0.0, and additional changes are implemented in PX-Backup to accommodate the newer Keycloak version. This provides a seamless experience with reduced number of vulnerabilities reported in the earlier versions. |
Known Issues (Errata)
Portworx is aware of the following issue, check future release notes for fixes on this issue:
Issue Number | Issue Description |
---|---|
PB-1840 | Issue: After upgrading from the earlier PX-Backup version to 2.0.1 with Keycloak or an external OIDC, sign in to OIDC fails. Workaround: Clear your browser cache, and then sign in. |
2.0.0
July 30, 2021
New features
PX-Backup includes the following new features and enhancements:
- MongoDB: replaced the existing etcd backup store with MongoDB to improve the scalability and performance of PX-Backup.
- PX-Backup Security: enables the infrastructure administrators and application owners to control the level of access to certain PX-Backup resources by setting governance policies and managing permissions on the platform. See PX-Backup Security for more information.
- Backup Timeline: provides a daily and monthly graphical view of backup activity and status. See PX-Backup Timeline for more information.
- Application grouping: provides an optimized view for PX-Backup applications, instead of individual resources. Improved scalability to handle large numbers of enterprise resources and namespaces. See Create a backup topic for the updated Applications UI.
- Guided onboarding: provides a step-by-step guided experience for configuring and managing PX-Backup.
Improvements
Pure Storage has upgraded or enhanced functionality in the following areas:
Improvement Number | Improvement Description |
---|---|
PB-1092 | The new PX-Backup Applications page UI allows users to choose resources at each namespace level, as needed. |
PB-1233 | IBM cloud is now available as a cloud provider in the PX-Backup Add Cloud Credentials page. |
PB-1539 | Reduced vulnerabilities in the updated PX-Backup base image. |
PB-1560 | The Default cloud credentials, which are public and accessible by all PX-Backup users, is removed from Cloud Credentials page. |
PB-1673 and PB-1748 | Various improvements around Restore progress bars |
PB-1688 | A warning message now appears on the PX-Backup Applications page when the Stork version is detected prior to 2.6.4. PX-Backup enables certain features (for example, custom resource selection), only if Kubernetes cluster is installed with Stork version 2.6.4 and above. |
Fixes
Pure Storage has fixed the following issues:
Issue Number | Issue Description |
---|---|
OC-569 | Cannot specify nodes while installing or enabling PX-Central using Helm. User Impact: Unable to select nodes using the nodeSelector flag while installing PX-Central using Helm. Resolution: While enabling or installing PX-Backup and monitoring service, you can select nodes using the nodeAffinityLabel flag. If the nodes are labelled with px/central key, then you can set the following helm parameter:--setnodeAffinityLabel=px/central |
OC-588 | PX-Central services are automatically configured with LoadBalancer and/or NodePort while deploying PX-Central or PX-Backup. User Impact: Cannot configure all PX-Central services explicitly as the ClusterIP . Resolution: While installing or upgrading PX-Central chart, you can set up the services according to your environment using the following command:--set service.pxBackupUIServiceType=ClusterIP,service.grafanaServiceType=ClusterIP,service.cortexNginxServiceType=ClusterIP |
OC-742 | When you install PX-Backup, the pxcentral-backend and px central-frontend pods do not start. User Impact: The px-backup chart fails on Kubernetes version 1.21. Resolution: The PX-Central installation now supports Kubernetes versions from 1.16 to 1.21. |
PB-1227 | PX-Backup should not prompt IKS users to manually update the cluster kubeconfig whenever the access token expires. User Impact: Users couldn’t use the IBM cluster kubeconfig when the access token expires. Resolution: PX-Backup now automatically refreshes the IBM cluster kubeconfig when users provide IBM credentials with an API key. |
PB-1228 | If a deleted namespace is in a terminating state, then PX-Backup Applications page displays its status as active. User Impact: Users may have seen the status of deleted namespaces were not updated in the PX-Backup Applications page. Resolution: The Applications page no longer displays a deleted namespace if it is in a terminating state. |
PB-1478 | When Keycloak is unreachable, PX-Backup does not display a proper error message. User Impact: When Keycloak was down, users saw an improper connection timeout message. Resolution: PX-Backup now displays a meaningful error message when Keycloak is down or the endpoint is unreachable. |
PB-1697 | While changing the PX-Backup token lifespan, the post-install hook pod fails and does not update the PX-Backup secrets with the required OIDC tokens. User Impact: PX-Backup installation fails because of intermittent Keycloak APIs failure. Resolution: Fixed with retries when Keycloak APIs fail. |
PB-1794 | In some cases, initializing time for the MySQL pod is longer than expected. During this prolonged initialization, the liveliness probe fails and Kubernetes deletes the pod. This causes MySQL to go into an inconsistent state. User Impact: MySQL pod never becomes healthy. Resolution: The initialDelaySeconds parameter is increased to minimize the initialization time for the MySQL pod. |
PB-1800 | An incorrect check for comparing includeResource in the backupschedule reconciler results in PX-Backup pod to dump failed comparison logs on the ApplicationBackupSchedule object. User Impact: The following message appears in the PX-Backup logs:time="2021-07-27T16:19:01Z" level=info msg="backupScheduleStatusUpdateReconciler: updated backup schedule CR mysql-csi-sch-46c06bb on cluster ocp-cluster" time="2021-07-27T16:19:02Z" level=debug msg="compareBackupSchedule: backup schedule mysql-csi-sch - orgID default IncludeResources differ Resolution: The incorrect comparison is resolved and PX-Backup pod does not display the messages in the log. |
Known Issues (Errata)
Portworx is aware of the following issues, check future release notes for fixes on these issues:
Issue Number | Issue Description |
---|---|
OC-748 | The px-backup-ui service will be deprecated in the future PX-Central versions. User Impact: No user impact in the PX-Central 2.0.0 version. It may occur in the future version. Recommendation: In the future PX-Central versions, use the px-central-ui service to access the UI. |
PD-889 | If a cluster contains a larger number of nodes, removing the cluster from PX-Backup UI and re-adding it displays a warning. User Impact: Users can view that the cluster is not added, but internally the cluster is added. Recommendation: Users can check the cluster details page if the cluster is added. If the cluster is still not added, try adding it again. |
PB-1308 | If the IP address of a server is pointing to a loopback address (for example, 127.0.0.1:6443), instead the actual IP address in the kubeconfig, adding an application cluster using this kubeconfig will fail. User Impact: Cannot add an application cluster when using the loopback IP address in kubeconfig. Recommendation: Replace the loopback IP address with the IP address of the Kubernetes master. |
PB-1382 | While installing PX-Backup 2.0.0 using Helm, the CustomResourceDefinition is deprecated warning appears. User Impact: PX-Backup installation fails because the apiextensions.k8s.io/v1 API version is not supported with the Kubernetes versions prior to 1.16. Recommendation: To install PX-Backup with the Kubernetes versions prior to 1.16, users should:
|
PB-1785 | For adding a backup location using the IBM cloud, the Secret Key and Access Key are mandatory. User Impact: Without the Access Key and Secret Key, users cannot add a backup location using the IBM cloud account. Recommendation: For adding a backup location using the IBM Cloud Object Store. |
1.2.4
May 10, 2021
Improvements
Pure Storage has upgraded or enhanced functionality in the following areas:
Improvement Number | Improvement Description |
---|---|
PB-1419 | PX-Backup now handles license activation for clusters deployed from the IBM Cloud Catalog and running in air-gapped mode or VPC private only network. |
1.2.3
April 2, 2021
Improvements
Pure Storage has upgraded or enhanced functionality in the following areas:
Improvement Number | Improvement Description |
---|---|
PB-1097 | Improved drop-downs in Safari. |
PB-1125 | If you delete multiple backups, the delete confirmation modal now displays the number of backups being deleted instead of a list of backups. |
PB-1149 | If you are an IBM Cloud Pay-As-You-Go or subscription account user, now you cannot import a different type of license. |
PB-1160 | The Protected Data field now displays a sum of sizes of all backups from all clusters, including deleted clusters. |
PB-1145 | PX-Backup now properly displays the start date of your IBM Cloud Pay-As-You-Go or subscription account license. |
Fixes
Pure Storage has fixed the following issues:
Issue Number | Issue Description |
---|---|
PB-1077 | Sometimes, PX-Backup needed more than one iteration to delete a backup, even if there were no dependent backups. User Impact: PX-Backup took a long time to delete a backup. Resolution: PX-Backup now deletes backups faster, as it places all backups in the same iteration. |
PB-1078 | On the Backups tab, selecting a date range had no effect. User Impact: PX-Backup always displayed all backups. Resolution: PX-Backup now displays backups taken during a specific date range. |
PB-1090 | Sometimes, when the number of resources being enumerated was very large, background workers took too much memory when they iterated over backups. User Impact: The PX-Backup containers were evicted. Resolution: Background workers no longer take too much memory when they iterate over backups. |
PB-1123 | When a user who didn’t have permission to create a Kubernetes namespace tried to create a backup schedule, the operation failed even when the namespace was already present. User Impact: Users who didn’t have permission to create a Kubernetes namespace could not create a backup schedule. Resolution: When a user who doesn’t have permission to create a Kubernetes namespace tries to create a backup schedule, and if the namespace already exists, then PX-Backup will create that backup schedule. If the namespace does not exist and the user does not have permission to create a namespace, then the operation will fail. |
1.2.2
Jan 26, 2021
New features
- PX-Backup now allows you to delete multiple backup and restore jobs.
- Users can recreate a backup or restore job from an existing job by duplicating a successful or failed backup and restore job.
- If you’re using a CSI driver and the original cluster is no longer available, you can now choose any other CSI cluster to delete your CSI backup.
- You can now specify the CSI snapshot class that PX-Backup will use to back up a CSI volume.
- PX-Backup now supports cross-cluster restores on clusters running Pure Service Orchestrator (PSO) v6.0.5
Improvements
Pure Storage has upgraded or enhanced functionality in the following areas:
Improvement Number | Improvement Description |
---|---|
PB-773 | PX-Backup now displays an improved error message when users choose the “Include any namespace created” option on a cluster running a version of Stork older than 2.5.0. |
PB-981 | PX-Backup now displays different icons for partially successful restores. These icons help to distinguish between partially successful restores and successful ones. |
PB-1050 | When the number of resources that are backed up is very large, the PX-Backup Details modal now displays a message indicating that resources are being loaded. |
Fixes
Pure Storage has fixed the following issues:
Issue Number | Issue Description |
---|---|
PB-1003 | The view JSON output for a backup schedule did not show all the successful backups taken by that backup schedule. User Impact: The users could not see all their successful backups. Resolution: The view JSON output for a backup schedule now shows all the successful backups thats were taken by that backup schedule. |
PB-1015 | Sometimes, when a cluster came back online, PX-backup did not update the status of that cluster immediately. User Impact: There was a delay in updating the cluster status during which users saw their online clusters marked as offline. Resolution: PX-Backup now correctly reflects the status of your cluster. |
PB-1036 | Users could not apply a new license. User Impact: PX-Backup displayed the following error message: “Can’t update license as current license type is invalid.” Resolution: Users can now apply new licenses. |
PB-1039 | When the number of namespaces being backed up was very large, PX-Backup took a long time to load all resources, and backups would fail with a timeout error. User Impact: PX-Backup marked the backup as failed. Resolution: PX-Backup no longer times out and marks backups as failed when the number of namespaces being backup up is very large. |
PB-1047 | On the Applications page, if the user selected a particular resource type to back up, PX-Backup enabled the Backup button before all the resources were loaded. User Impact: Sometimes, PX-Backup backed up only a subset of the resources the user has selected. Resolution: PX-Backup now enables the Backup button only after it loads all resources, and backs up all resources. |
PB-1056 | Backup jobs became stuck in the “In progress” state when the application cluster on which you triggered the backup has been shut down or terminated. User Impact: Users saw these backup jobs sit in the “In progress” state in the PX-Backup UI and never converge to the “Failed” state. Resolution: PX-Backup now correctly marks backup jobs as “Failed” when the application cluster on which you triggered the backup has been shut down or terminated. |
PB-1063 | If PX-Backup failed to create a backup location, the objects created to validate the cloud credentials were not cleaned up. User Impact: If PX-Backup failed to create a backup location, the objects created to validate the cloud credentials were not cleaned up. Resolution: When PX-Backup fails to create a backup location, it now removes all objects created to validate the cloud credentials. |
PB-1068 | PX-Backup did not verify the license immediately after the PX-Backup pod was restarted. User Impact: If the license was expired and the users restarted the PX-Backup pod, there was a ten minutes period during which users could create backups Resolution: PX-Backup now verifies the license immediately after the PX-Backup pod is restarted. If the license is expired, the backups fail, and PX-Backup displays an error saying that the license is expired. |
PB-1069 | Sometimes, PX-Backup marked a backup job as “Done” even if a volume backup was still in progress. User Impact: Sometimes, users saw their jobs being marked as “Done” even if a volume backup was still in progress Resolution: PX-Backup now marks a backup job as “Done” only after all volume backups are successfully completed. |
1.2.1
Jan 4, 2021
Fixes
Pure Storage has fixed the following issues:
Issue Number | Issue Description |
---|---|
PB-1034 | The users could not use a license file to activate a license. User Impact: PX-Backup displayed the following error message: “no license provided for activation.” Resolution: The users can now use a license file to activate a license. |
1.2.0
Dec 3, 2020
New features
- Introducing usage-based pricing for remote cluster nodes.
- PX-Backup now supports generic CSI driver backup and restore.
- Introducing cluster-level aggregated metrics for backup and restore with Prometheus metrics and Grafana dashboards
Improvements
Pure Storage has upgraded or enhanced functionality in the following areas:
Improvement Number | Improvement Description |
---|---|
PB-553 | On the Restores page, PX-Backup now displays a different icon for partially successful restores. This icon helps to distinguish between a partially successful restore and a successful one. |
PB-894 | On the All Backups page, PX-Backup now displays the name of the cluster for each backup. |
PB-932 | On the Schedules page, you can now hover over a paused backup schedule to see the reason for that backup being paused. |
PB-934 | On the Applications page, PX-Backup now properly displays namespaces with long names in the namespace drop-down list. |
PB-948 | PX-Backup now displays the date and the time when the next backup will run in the modal containing details about your backup schedule. |
PB-946 | PX-Backup now uses the same format to display dates for backups, restores, and backup schedules. |
PB-947 | On the Scheduled Backup Details pane, the NEXT SCHEDULED BACKUP ON field now shows the date and the time when the next backup will run. When a backup is running, the NEXT SCHEDULED BACKUP ON field shows In Progress. |
PB-950 | If you select the name of a backup, restore, or backup schedule from the table view, PX-Backup now displays a modal containing details about your backup, restore, or backup schedule. |
PB-958 | On the Applications page, the list of resource types is now sorted alphabetically. |
PB-966 | On the Schedules page, PX-Backup now displays the namespaces included in a backup. |
Fixes
Pure Storage has fixed the following issues:
Issue Number | Issue Description |
---|---|
PB-659 | PX-Backup did not notify the users when their license was about to expire. User Impact: The users would know about expired licenses only when backups and restores started failing due to expired licenses. Resolution: PX-Backup now displays a warning seven days before license expiration. |
PB-862 | If the OIDC server was not reachable during startup, PX-Backup failed to start User Impact: PX-Backup did not start. Resolution: If the OIDC server is not reachable during startup, PX-Backup now starts and tries to connect to the OIDC server before a gRPC call is performed. |
PB-892 | PX-Backup incorrectly reflected the size of your AWS backup. User Impact: PX-Backup displayed “B” instead of “GiB” Resolution: PX-Backup now accurately displays the size of your AWS backup. |
1.1.1
Nov 2, 2020
Improvements
Portworx, Inc. has upgraded or enhanced functionality in the following areas:
Improvement Number | Improvement Description |
---|---|
PB-868 | On the Application page, you can now select the Backup button to perform a new backup operation, even if not all resources are loaded yet. |
PB-809 | When an API call times out, PX-Backup now displays a more descriptive error message containing the full URL of the API call. |
PB-871 | PX-Backup now displays an error message when a user that does not have adequate permissions to add a new cluster to PX-Central tries to add one. |
PB-901 | On the Add Backup Location page, the Endpoint field can now have a maximum of 512 characters. |
Fixes
Portworx, Inc. has fixed the following issues:
Issue Number | Issue Description |
---|---|
PB-739 | If you used Swagger to query a backupLocation , PX-Backup returned an empty response. User Impact: Swagger displayed the following error message: “Method Not Allowed /v1/backuplocation”. Resolution: If you use Swagger to query a backuplocation , PX-Backup no longer returns an empty response. |
PB-776 | If a user wished to restore a backup, PX-Backup selected all namespaces included in that backup for restore. User impact: A restore could cause unwanted configuration changes on the destination cluster. Resolution: PX-Backup no longer selects all namespaces for restore. The users choose which namespaces they want to restore. |
PB-856 | If you logged in for the first time and no clusters were added to PX-Backup, the dashboard indicated that PX-Backup must still load the stats. User impact: The stats section of the dashboard displayed three dots. Resolution: The dashboard now clearly shows that no clusters are added to PX-Backup, by setting all stats to zero. |
PB-875 | Backups sometimes became orphaned, losing the association with their objects in a data store. User impact: When the user tried to delete a backup that depends on an orphaned backup, that backup became stuck in the “Delete Pending state”. Resolution: PX-Backup now deletes scheduled backups that become orphaned. If a manual backup becomes orphaned, then you must remove its corresponding object from the data store. |
PB-857 | If two users added the same cluster to PX-Backup, and one user doesn’t have adequate permissions to list the nodes in the cluster, then the status of the cluster is incorrectly reflected in PX-Backup. User Impact: For both users, the status of the cluster changed continuously from “Active” to “Inactive”. Resolution: PX-Backup now accurately displays the status of the cluster. |
PB-655 | PX-Backup failed to create a backuplocation in the AWS us-west-2 region when the user provided the default endpoint (s3.amazonaws.com ). User Impact: PX-Backup displayed the following error message: “backup location [awsl1] creation failed as provided cloud credential [awscc] is not valid: cloud credential [awscc] doesn’t have permission to upload object: BucketRegionError: incorrect region, the bucket is not in ‘us-east-2’ region at endpoint ‘s3.amazonaws.com’ status code: 301, request id: , host id:” Resolution: PX-Backup now creates a backuplocation in the AWS us-west-2 region, even if the user provides the default endpoint (s3.amazonaws.com ) |
1.1.0
Sep 28, 2020
New features
- If you add a new cluster using the CLI or API, PX-Backup now displays your cluster in the UI.
- Added a separate Lighthouse view.
- The new PX-Backup dashboard provides insights into your protected applications. You can view the amount of data backed up, and the policies enforced both at the individual cluster level and the multi-cluster level.
- PX-Backup now features resource-level backups, allowing you to perform granular backup operations by resource type and also at the individual resource level.
- PX-Backup now features selective restores, allowing you to selectively restore specific resource types or resources from any selected backup.
- Introducing default backup policies: administrators can now use wildcards to specify backup policies. PX-Backup will add all newly created namespaces to that backup schedule, without requiring a policy update.
- Administrators can now share the default cloud account and backup location with other users.
- Added additional metrics for backups, including the size of backups per PVC, namespace, and cluster.
- To help improve user experience, PX-Backup now uses telemetry to collect information about your use cases, backup metrics, and deployment environments.
- PX-Backup licenses are node-based, and you can check the node count when you import a license.
Improvements
Improvement Number | Improvement Description |
---|---|
PB-783 | PX-Backup now validates the bucket name field in the Add Backup Location view. |
PB-762 | When a backup schedule is in the delete pending state, PX-Backup no longer displays the remove, suspend, or edit options. |
PB-837 | When a backup location is in the delete pending state, PX-Backup no longer displays the remove option. |
PB-640 | The Backups view now features a progress bar indicating the progress of your backup operations. |
PB-682 | The Backup Rules page now includes a help message explaining pre and post backup rules. |
PB-680 | Persistent volumes no longer appear in the Restore Backup view. |
PB-706 | Improved validation rules for the field that specifies the number of scheduled backups that PX-Backup retains. |
PB-699 | The Restores view now features a progress bar indicating the progress of your restore operations. |
PB-671 | PX-Backup now displays the backup size for each namespace in the Restore Backup view. |
PB-478 | When a backup is in the delete or delete pending state, PX-Backup no longer displays the View json and Show Details options. |
PB-302 | PX-Backup now automatically validates backup locations when the users add them. |
PB-634 | Users can now filter resources by resource type. |
PB-768 | Users can now delete a resource without being prompted for the name of that resource. |
PB-636 | Improved clarity around the OrgID field in the Add License view. |
PB-645 | In the Edit Backup Schedule View, you can now use a navigation link to go to the Schedule Policy view. |
PB-710 | Added a tooltip showing whether a backup schedule is being deleted. |
PB-509 | When your Keycloak token expires, PX-Backup now redirects you to the login page. |
PB-664 | PX-Backup now displays the status of a cluster as Inactive , even if the cluster has been deleted or is not reachable. |
PB-712 | Every time you update the cluster configuration, PX-Backup validates whether the cluster is accessible. |
PB-637 | If your license expires, PX-Backup pauses all scheduled backups until you apply a new license. |
PB-654 | The users are no longer required to provide the org name when they generate new license files. |
PB-745 | The Backup Rules view now features a new Container field that allows the users to specify the container to which PX-Backup will apply the rule. |
PB-831 | PX-Backup now displays the resources in alphabetical order in the Create Backup view. |
Fixes
Portworx, Inc. has fixed the following issues:
Issue Number | Issue Description |
---|---|
PB-629 | Users were unable to log out and log in as a different user. User impact: They were seeing an error message saying “You are already authenticated as different user |
PB-686 | If you provide an HTTP endpoint in the backup location field, sync backup fails. User impact: PX-Backup displays the following error: “Access Denied.” Resolution: Sync backup now works, even if you provide an HTTP endpoint. |
PB-608 | Middleware not able to establish a connection with backup because the grpc connection was not closed. User impact: Because of this issue, PX backup was marked as offline. Resolution: Users will no longer see PX-backup marked as offline due to the middleware not being able to establish a connection with PX-Backup. |
PB-664 | If a backup schedule is associated with an inactive cluster, users can not remove the cluster from PX-Backup. User impact: PX-Backup displays an error message saying that the user can not delete the cluster. Resolution: The users can now select the inactive cluster and perform all operations except triggering a new backup. |
PB-744 | When the users created a large number of backups, and each backup contained a large number of resources, the All backups view did not show any backups. User impact: They could not see their backups in the All backups view. Resolution: PX-Backup always shows all backups in the All backups view, even if the users create a large number of backups and resources. |
1.0.2
July 28, 2020
Improvements
Portworx, Inc. has upgraded or enhanced functionality in the following areas:
Improvement Number | Improvement Description |
---|---|
PB-621 | Generic CRD support: PX-Backup now shows CRs in the application view. You must use Stork 2.4.3 or greater on the application cluster. |
PB-574 | Added support for backing up namespace quotas |
PB-573 | Added support for the kubectl oidc authenticator |
PB-565 | Provided an option to copy the json output from the Inspect Data pane |
PB-539 | PX-Backup now displays the orgID in the user’s profile page |
PB-464 | The scheduled backups settings now use a 12-hour clock |
PB-609 | The tooltip now shows the reason for PX-Backup being marked offline when you hover over it |
PB-584 | The restore view now features a progress bar |
PB-576 | The backup view now features a progress bar |
PB-575 | Added a help message to explain the Path / Bucket field in the backup location screen |
PB-572 | PX-Backup now reads the OIDC admin secret into a user-provided namespace instead of the PX-Backup namespace. |
Fixes
Portworx, Inc. has fixed the following issues:
Issue Number | Issue Description |
---|---|
PB-627 | Backup location, schedules, pre, and post rule dropdowns showed only 10 entries, even if there were more. User impact: If they had more than 10 entries, users couldn’t access them from the dropdowns. Resolution: PX-Backup now shows all results in these dropdowns. |
PB-623 | Users were unable to delete restore jobs that were in the pending state. Resolution: Users can now delete pending restore jobs. |
PB-610 | Due to a race condition between the schedule delete and reconciler status updates, PX-Backup did not delete backup schedules when prompted to. Resolution: PX-Backup now properly deletes backup schedules. |
PB-608 | Middleware not able to establish a connection with backup because the grpc connection was not closed. User impact: Because of this issue, PX backup was marked as offline. Resolution: Users will no longer see PX-backup marked as offline due to the middleware not being able to establish a connection with PX-Backup. |
PB-599 | Stork continuously retried to update the backup/restore resources when PX-Backup marked a job as failed. User impact: In some cases, Stork would eventually mark the backup CR as successful, but PX-backup would continue to show it as failed. Resolution: PX-backup now accurately reflects the backup CR’s status. |
PB-590 | Cloud credential information was displayed in plain text in the logs and in the View JSON option. Resolution: PX-Backup no longer displays credential information in these places. |
PB-579 | Restore jobs sometimes became stuck in the pending state. User impact: Users would see these restore jobs sit in the pending state in the PX-Backup UI and never converge to a failed state. Resolution: If the restore job is stuck in a pending state, it will eventually be marked as failed after the timeout period. |
PB-578 | Backup entries were not deleted from the PX-Backup UI when backup sync was in progress and the backup location was deleted User impact: Users would see backup entries from a backup location that was removed from PX-backup UI. Resolution: PX-Backup now deletes these backup entries. |
PB-569 | “Successfully” is no longer misspelled in the Restore status dialog. |
PB-552 | PX-Backup failed to indicate that users must have admin privileges when adding a Portworx cluster. User impact: Users may not have known why they couldn’t add a Portworx cluster. Resolution: In the Portworx endpoint section, a message now indicates that admin account privileges are needed for to add a Portworx cluster for monitoring. |
PB-541 | Clusters with PX-Backup disabled were listed on the dropdown in the PX-Backup dashboard User impact: Users may have been confused by these erroneous listings Resolution: PX-Backup no longer lists clusters with PX-Backup disabled. |
PB-384 | PX-Backup picked up the existing token when users signed out and attempted to sign in User impact: Users would be directly signed-in when they attempted to log back in and could not switch users if desired. Resolution: PX-Backup now redirects users to the sign-in page after logging out. |
PB-607 | It was possible to enter decimal numbers into the schedule dialog. Resolution: The schedule dialog no longer accepts decimal numbers as input. |
1.0.1
June 5, 2020
Improvements
Portworx, Inc. has upgraded or enhanced functionality in the following areas:
Improvement Number | Improvement Description |
---|---|
PB-547 | PX-Backup now allows more than 12 backups to be retained while creating a schedule. |
PB-515 | Users can now add a cluster to Lighthouse and and edit it independently from PX-Backup. |
PB-389 | The App View page now includes a refresh button, allowing you refresh the list. |
PB-485 | The credential settings page now includes info icons explaining what should be entered into the input fields. |
PB-480 | The Remove button no longer appears when there is no entry in the pod selector during Rule creation. |
PB-479 | When a backup deletion is pending, PX-Backup no longer shows a restore option. |
PB-455 | An improved error message now displays when Stork is not installed on the application cluster. |
PB-453 | Selections on the namespace selection list now persist when you switch between tabs. |
PB-451 | When adding a Google cloud account, you can now upload your json key using the file browser. |
PB-444 | Improved clarity around options for pasting or uploading your kubeconfig on the Add Cluster page. |
PB-435 | A new warning message now indicates that any backups that belong to a deleted backuplocation will also be deleted. |
PB-507 | A cluster with a status of Inactive is now highlighted when the cluster is down for improved visibility. |
PB-500 | The field labels in the Add cloud account page have been improved. |
Fixes
Portworx, Inc. has fixed the following issues:
Issue Number | Issue Description |
---|---|
PB-534 | Clusters could be added to PX-Backup using credentials that did not have adequate permissions. User Impact: PX-Backup would not be able to display data or perform operations properly. Resolution: Cluster addition now fails if credentials do not have all of the required permissions. |
PB-519 | It was possible to attempt to add a backup location before adding a cloud account. User Impact: Despite seeing the + Add option, users would not be able to add a backup location. Resolution: Users must now add a cloud account before seeing the option to add a backup location. |
PB-400 | When editing a backup schedule, the existing pre-exec and post-exec rules did not appear in the dialog box. User Impact: Users would have to reselect their pre-exec and post-exec rules when they edited an existing backup schedule. Resolution: PX-Backup now lists the existing pre-exec and post-exec rules when a user edits a backup schedule. |
PB-307 | Restores from subsequent backups stored on different buckets from the original that both involve the same namespace failed. User Impact: If users created multiple backups and buckets, restores from subsequent backups stored on different buckets from the original that both involve the same namespace failed. Resolution: PX-Backup now properly takes the initial backup for incremental backups, even when they share a namespace with other backups using different buckets. |
PB-499 | When PX-Backup failed to create a backup, the reason for the failure did not propagate to the UI correctly, showing only the creation failure message. User Impact: Due to the stuck message, users may have been unable to see the real reason for a backup deletion failure. Resolution: Creation failure messages no longer stick, allowing users to see more recent messages. |
PB-493 | The PX-Backup UI did not update to reflect invalid underlying backup object statuses. User Impact: If a backup object was deleted, users may still have seen the backup appear as valid on the PX-Backup UI. Resolution: PX-Backup now properly displays the invalid backup status. |
PB-468 | After the trial license expired, there was no way to let the user know that they had more clusters online on PX-Backup than the license supported until the backup failed. User Impact: Users would not know they have more cluster than the license supports until their backup or restore failed. Resolution: PX-Backup now displays a message on the home screen to let the users know that they need to remove some clusters if they have more clusters than their license supports. |
PB-395 | The PX-Backup app view took a long time to load when multiple namespaces were selected. User Impact: Users selecting a large number of namespaces in the app view would either not see, or have to wait a long time for resources to appear Resolution: PX-Backup now loads resources for multiple namespaces faster. |
PB-486 | When a user deleted a backup, PX-Backup did not update the reason. User Impact: Users may have seen confusing information about their backup status during deletion Resolution: PX-Backup now shows an appropriate message indicating volumes and resources are being deleted. |
PB-477 | The enter key did not work in search box on the All backups page. User Impact: Users couldn’t filter their results on the All backups page using the search box. Resolution: The search box now filters results when a user inputs a term and presses enter . |
1.0
April 30, 2020
New features
Announcing PX-Backup, a Kubernetes backup solution that allows you to back up and restore applications and their data across multiple clusters.
PX-Backup includes:
- Point-and-click recovery for any Kubernetes app—stateless or stateful
- Fast recovery for applications, including restoring their configuration and data
- Pod, tag, and namespace granularity for any backup
- Policy-driven scheduled backups
- Continuous backups across multiple clouds and global data centers
- Complete protection for applications, configurations, and data
- Application-consistent backup and restore
- Backup and recover cloud volumes from AWS, Azure, and GCP, even if you are not using PX-Store
- Built-in multi-tenancy for enterprises via industry-standard OIDC integration