The following known issues have been discovered through rigorous testing and will help you understand some problems you might encounter in this release. The list of issues below pertains to this release of Data Recovery only.
-- ///*** end of content area ***/// -->
- Deduplication Store May Remain Locked After Data Recovery Exits Unexpectedly
The deduplication store reduces the risk of data corruption by using a locking mechanism. This mechanism helps prevent multiple instances of Data Recovery from writing to the same deduplication store. If Data Recovery has locked the deduplication store as part of the backup process, but then exits unexpectedly, the deduplication store is left in the locked state. Upon restarting Data Recovery and attempting to complete tasks, the following may occur:
- The attempt to manually start the backup fails and the following error occurs:
The destination is not available for a backup job. Please ensure the destination has been formatted and there is enough space.
- The automatic backup fails and the following error occurs:
Trouble writing files, error -1020 (Sharing violation).
If Data Recovery has recently stopped unexpectedly and the error occurs, check the backup appliance for the store.lck folder, which is found in
/VMwareDataRecovery/BackupStore/store.lck. If this folder exists and you are certain there are no other instances of Data Recovery using this deduplication store, delete the lock folder. If you delete this folder and other instances of Data Recovery are using the deduplication store, corruption may occur. After deleting the lock folder, Data Recovery can complete operations as expected.
- If the Deduplication Store Is Full No Actions May Be Taken For 12 Hours
Normally, if the deduplication store becomes full, Data Recovery runs the retention policy. If the retention policy has already run within the last 12 hours, the policy is not run again until 12 hours since the last time retention policy was run. Adding more storage or modifying the retention policy so fewer backups are kept may resolve the lack of space in the deduplication store. These changes may make further backups possible, but Data Recovery does not attempt to complete backups for at least 12 hours, even if there are jobs scheduled to run before the 12 hours passes. To resolve this issue, wait for the 12 hour window to pass.
- Restore Wizard Does not Enforce Valid Datastore Selections
Using the restore wizard, it is possible to specify datastore locations to which to restore virtual machines where the datastore is not valid. For example, when:
- A datastore is renamed or deleted, the old information persists, so an outdated name or non-existent store can be selected.
- A virtual machine is to be restored to a cluster, datastores that are not on shared storage may be selected.
To avoid this issue, be certain the datastore selected in the wizard is valid.
- Restoring Virtual Disks Never Powers on a Virtual Machine
It is possible to restore a backed up virtual disk, while not restoring the virtual machine on which the disk is installed, leaving it in its present state. The restore process includes an option to power on the virtual machine whose disk is restored. If only the disk is being restored, but not the virtual machine, the virtual machine is not powered on, regardless of the configured settings. To resolve this issue, power on the virtual machine after restoring the disk.
- Restored Virtual Machines Are not Connected As Expected
When restoring Virtual Machines, they are configured with the same network as at the time of the backup. If a virtual machine was backed up while connected to a network, and that network is no longer available, when the virtual machine is restored it cannot connect to the network because its network configuration is invalid. To resolve this issue, use vSphere client to edit its settings and change its network configuration so that it is connected to a valid network.
- Manual Bring To Compliance Function Limited to Eight Virtual Machines
By right-clicking a backup job, it is possible to select the Bring to Compliance option. Doing so automatically backs up the virtual machines that are covered by the policy and that have changes that have not been backed up. This function is limited to backing up no more than eight virtual machines at a time. This limitation applies to any case where more than eight virtual machines are queued for backup. This could occur because more than eight were affected by the selection of a particular backup job or it could happen because of an additional Bring to Compliance operation that brings the total number over eight. If more than eight virtual machines are in the Bring to Compliance queue, only eight of the virtual machines are backed up. To avoid this issue, do not attempt to manually bring more than eight virtual machines in to compliance concurrently.
- Virtual Machine Templates Are not Displayed for Backup
Data Recovery does not support virtual machine template backup. Therefore, these objects are not shown in the Data Recovery user interface.
- Restore Points May not Be Displayed in the vSphere Client
In rare cases, the Restore tab may not show any restore points, even though valid restore points exist. You can confirm that valid restore points exist since they are displayed in the Restore Wizard. To resolve this issue, select a different tab and then select the Restore tab. If the problem persists, restart the vSphere Client.
- Removed Hosts and Clusters not Updated in vSphere Client as Expected
Data Recovery displays available hosts and clusters in the vSphere Client user interface. If a host or cluster is removed from vSphere, renamed, and re-added to vSphere, both the new and old instances are displayed in the Data Recovery plug-in. To resolve this issue, restart the backup appliance, thereby refreshing Data Recovery's information about the inventory.
- The Backup Appliance Must not Be put in a vApp Folder
Data Recovery is unable to find the backup appliance if it is place in a vApp folder. To resolve this issue, move the backup appliance in a different location.
- The Backup Appliance May Incorrectly Boot from Hotadded Disks
If the backup appliance fails while hotadded disks are attached, the backup appliance may attempt to boot off of a hotadded disk instead of the disk with the backup appliance's operating system. If you encounter this issue, contact VMware support.
- Deduplication Store Activity May Trigger CPU Usage Warnings
The tasks completed by the deduplication store are CPU intensive. As a result, the CPU usage warnings may be triggered, although these warnings typically do not indicate a problem. To avoid this issue, reconfigure CPU usage warnings for the backup appliance.
- Newly Added SCSI Disks Are not Visible If Added While a Task Is Running
If a SCSI disk is added to the backup appliance while the backup appliance is completing a task such as a backup or restore, the new disk is not displayed, so it cannot be configured or used. To resolve this issue, wait for the running task to complete, after which the disk becomes visible and configurable.
- Unmounted Disks Do not Appear as Expected
The backup appliance shows both mounted and unmounted virtual disks. Unmounted virtual disks have a device path such as /dev/sdx and mounted virtual disks have a SCSI address such as /SCSI-0:1. In some cases, if the mounted virtual disk is unmounted, the disk with the SCSI address path such as /SCSI-0:1 is removed but the unmounted disk is not then shown with the device path such as /dev/sdx. To resolve this issue, reboot the backup appliance.
- Added Network Shares May Have No Name
If a network share is added, it may show up without a valid name. If this occurs, unmount and re-add the share until the name appears as expected.
- Restored Virtual Machines Attempt to Boot from IDE Drives
Virtual machines with both IDE and SCSI drives attempt to boot from an IDE drive after being restored. This occurs regardless of whether the operating system is on the IDE drive or not. To resolve this issue, change the boot order in the virtual machine BIOS so the system boots up using the appropriate disk.
- Users Are Asked For Credentials Multiple Times During Backup Appliance Setup
Users are asked for a username and password for a vCenter administrator account twice when first deploying the backup appliance. When first connecting to the appliance, the credentials are required to authenticate to vCenter. After the user has authenticated with vCenter, the appliance UI appears and the Getting Started wizard starts. The Getting Started wizard also requests vCenter credentials for use during backups. Once the backup appliance has stored valid credentials, the appliance automatically uses these credentials to authenticate to vSphere.
- No Restore Points May Be Displayed During Some Data Recovery Operations
During some operations, such as rebuilding the deduplication catalog or completing an integrity check after a first integrity has failed, the Data Recovery user interface says that no restore points exist. This may give the impression that any previously created restore points have been permanently deleted. In actuality, there are no restore points available while Data Recovery is completing certain operations, but restore points will be available again after operations are finished. In rare cases, valid restore points are not redisplayed after operations are completed. If this occurs, restart the backup appliance.
- Logs May Appear to Indicate Incorrect Backup Types
Data Recovery reduces the amount of time and space that backups require by backing up only parts of virtual machine disks that have changed since the last backup. This process is called an incremental backup. Incremental backups are only possible with hardware version 7 virtual machines because such machines track which parts have changed. Hardware version 4 virtual machines do not support change tracking, so theirs are always full backups. For hardware version 7 virtual machines, even during the first backup, Data Recovery determines which areas of the disk store data and copies only those. Therefore, even when Data Recovery is backing up an entire new hardware version 7 virtual machine for the first time, only the parts of the disk containing data are read, so those backups are always listed as incremental.
Data Recovery supports full backup of entire virtual machines as well as incremental backup of only those parts that require backup. Incremental backup is only supported with more recent files. When only incremental changes have occurred on older virtual machines, users may expect an incremental backup to occur, but because those virtual machines do not support incremental backup, a full backup is done. Conversely, when a newer virtual machine is being backed up for the first time, so the entire virtual machine must be backed up, users may expect to see a full backup. In fact, an incremental backup is occurring, but the increment is the entire virtual machine, rather than a subset.
- Non-English Characters Are not Supported with Network Shares
If a network share name, user name, or password contain non-English characters, attempts to add the network share fail. To avoid this issue, use English characters for share names, user names, and passwords.
- Option to Unhide Detailed Backup Information is Unavailable
Data Recovery provides detailed information about each entry in the Backup and Reports tabs. Users can hide the detailed information. In the Japanese and Chinese versions of Data Recovery, after details about backups are hidden, there is no option to restore the display of this information. To resolve this issue, restart the vSphere client.
- Network Destinations Must Be specified Using IP Addresses
Using DNS-resolved names to specify network destinations is not supported with Data Recovery. When adding network destinations using Configure > Destinations, enter an IP address. For example, use \\192.0.2.12\share rather than \\example\share.
- Integrity Checks Do not Provide Accurate Progress Information
When an integrity check is running, a progress indicator is displayed. This progress indicator does not accurately reflect activity. When checking integrity, wait for the process to complete, at which time the progress bar changes from 0% to 100%.
- Virtual Machines Are Not Restored to Folders That Have Been Deleted
After backing up virtual machines, users can delete the folders that contained those virtual machines, and then restore the virtual machine backups. In such a case, Data Recovery restores the virtual machines to the host's root folder, rather than recreating the original directory structure that contained the virtual machines, as users may expect. To resolve this issue, create the desired directory structure and move the virtual machines to the appropriate location. This issue does not apply to virtual machines in resource pools.
- Restoring Virtual Machines With Thin Disks May Incorrectly Fill the Disk
The storage space saved by using thin disks results from only consuming space for those blocks that contain data. Hardware version 4 virtual disks do not support tracking which blocks actually contain data and which have yet to be used. As a result, when restoring a hardware version 4 thin virtual disk, the entire user defined capacity must be restored to ensure no valid blocks are omitted. This effectively converts a thin disk to a thick disk. Deleting the disk in question before restoring the virtual machine avoids this issue.
- Selecting Multiple Appliances to Back up One Virtual Machine May Create Problems
It is possible for users to configure multiple backup appliances to back up the same virtual machine. This may produce unexpected results and is not supported with this release. Because Data Recovery does not check to ensure this configuration has not been established, users must ensure they do not create such a condition.
- Automatically Powered On Virtual Machines May not Reflect Correct Power State
Data Recovery includes an option to power on virtual machines after restoring from a backup. After such a restore, the Data Recovery inventory may incorrectly indicate that the restored virtual machine is powered off. To see accurate information about the restored virtual machine's power state, check its status in the vSphere Client's Hosts and Cluster view.
- Backup Appliance May Be Unavailable If a vCenter Server Has More Than 500 Virtual Machines
If a vCenter Server is managing more than 500 virtual machines, then Data Recovery may be unable to connect to the backup appliance, preventing the administration and potentially the functioning of Data Recovery. This does not impact any other vSphere functions, so other services continue to be available. To resolve this issue, manage the additional virtual machines with other instances of vCenter Server so there are no more than 500 virtual machines managed by each server.
- Deleted Virtual Machines May not Restore to Expected Folder Location
A virtual machine can be deleted from the inventory and then restored using Data Recovery. If a virtual machine was in a folder in the VMs and Templates view, the virtual machine is restored to the root level of the ESX server for that view.
To resolve this issue, users may choose to manually move restored virtual machines to their intended locations.
- Data Recovery Does not Simultaneously Back up Multiple Virtual Machines in a Single vApp
Multiple virtual machines can be packaged together in solutions called vApps. Data Recovery simultaneously creates restore points for up to eight virtual machines, but Data Recovery does not back up more than one virtual machine in a vApp at a time. Single virtual machines in different vApps and stand alone virtual machines can all be backed up simultaneously, but because multiple virtual machines in a single vApp are not backed up simultaneously, it is possible for backup jobs to take longer than expected to complete. Aside from increased back up times, this behavior does not negatively affect Data Recovery.
- Users Are not Notified If the Deduplication Store Is Full
Data Recovery does not provide warnings if the deduplication store is running low on space or if no space remains. Check regularly to ensures sufficient space is available and add more storage or modify the backup retention policy, as required.