Skip to main content
Druva Documentation

Known issues

This topic provides a list of known issues.

General Issues

Issue Description



While creating new administrators, Phoenix does not display the drop-down list of the available organization after selecting one organization.


Move the cursor on the screen and again click on the Organizations option.


After a group administrator deletes a backup set, the group administrator cannot view the jobs of the deleted backup set on the Jobs page. However, the cloud and the organization administrators can view all the jobs of the deleted backup set. 


The Manage Activation Tokens page does not list the complete string of the old activation tokens. Phoenix does not allow you to copy the older tokens and therefore, disables the radio button for these old tokens.


The search filter on the Jobs page does not display the jobs for the deleted backup set.


The Backup Content and Backup Policy fields in the Backup Activity Report show "NA" values for the jobs that are a part of the job listing pages not traversed after the Phoenix Cloud is upgraded to version 4.7. 


Once traversed correct policy and backup content will be shown in the Backup Activity Report.


After you upgrade the Phoenix agent to version 4.7, Phoenix does not display alert details for the alerts generated before Phoenix Cloud Push 4.7.

When you click View details, Phoenix displays the following message:

Details are not available for the old alerts.


If the server is renamed after a job has run for the server, the job activity reports display the new server name for that job. However, the Jobs page displays the old server name for the corresponding job.


Phoenix does not generate an audit trail for a canceled job.


Phoenix does not generate audit trails for jobs that update administrator names and time zones.


The child settings for entities are shown up even if the parent setting is disabled.


Phoenix generates audit trails for operations where the administrators save entities without editing them.


Cannot upgrade Phoenix client with a version prior to 4.7.0 without configuring the server on which the client is installed.


Ensure that you configure the activated server before you upgrade the client installed on it.


On the Summary tab of the job details dialog, the Error Message field shows repetitive text for a few jobs. 


The Unicode characters in the fully qualified domain name (FQDN) of a server are displayed as question marks (?) in the Phoenix Management Console.


A group administrator who monitors the backup and restore jobs for the NAS devices, cannot view the Request Job logs button on the Detailed Logs tab of the Jobs Details page.

ROBO-13776 If an API call is made to the Druva Cloud Platform using a tool that parses special characters (specifically the % character) in the call as-is without converting it to its UTF-8 value, the Druva Cloud Platform returns the call with an internal server error. 

File Server Issues

Issue Description

If you configure Linux 6.3 servers to follow the default backup policy, backup is successful with the following error:

FileAgent: Backup failed for the file /cgroup/cpuacct/cgroup.event_control in fileset root


Exclude the cgroup folder from the backup. To know how to exclude folders, see Update backup content of a file server backup policy

ROBO-2025 If the File server is not reachable during a scheduled backup, the Jobs page does not display the entry.
  • ROBO-3496
  • ROBO-7980

File server will not backup files with names, not in the Unicode format.


Full backup on Ubuntu 16.04 (64bit) OS is successful with errors.


During backup, exclude all the files from the following path as they do not have read permission.

  • /var/lib/lxcfs
  • /run/lxcfs
  • /run/systemd/system


  • If the script execution time limit is not specified and the backup window expires during the pre-backup script execution, the pre- and post-backup script statuses are not updated on the Job Details page.
  • If the backup window expires during the post-backup script execution and then the post-backup script execution is killed, the pre-backup script status gets marked on the Job Details page but not the post-backup script status.
ROBO-7445 Backup fails when a configured File server is reactivated from the command line.
ROBO-9535 Phoenix agent can back up data from a folder even if access to the folder is restricted for the user account on the server under which the Phoenix Service is installed and running.

NAS Issues

Issue Description


Phoenix fails to perform back up from a share if the share path entered as the share name also contains folder names stored on the NAS share.


While adding a share to a NAS device on Phoenix Management Console, enter only the share path as the share name. Do not include file or folder path in the share path.


If the NFS share name or path contains a backward slash (\), the native operating system treats it as a part of the share name instead of a separator that indicates the folder hierarchy. For example, an NFS share name or path is entered as \hr\vol\ny, OS treats hr\vol\ny as a single share name hrvolny instead of recognizing the folder hierarchy. For example, the native OS treats the share path \hr\vol\ny as a single name hrvolny and does not recognize the folder hierarchy.

Similarly, if the backup folder paths in the backup sets attached to NFS shares contain backward slashes, the native OS treats the entire path as a single text string.   


Replace the backward slash with a forward slash in the share name and in the backup sets attached to NFS shares. For NFS shares, ensure the share name and folder paths in the backup sets contain forward slashes (/).


While registering a NAS device, Phoenix does not validate the device credentials and registers the device. Later, when you register a share under the device, Phoenix displays the following error message.

Failed to add NAS share as failed to validate NAS credentials on the below NAS proxies due to the mentioned respective reasons.

- Credential validation failed for the NAS proxies.
(<NAS device name>)

This may happen if the NAS device was registered with incorrect credentials.


Update the device credentials and again try to add the share. See Manage your NAS device for more information. 


If the user access permissions for files and folders on the restore location are not the same as those on the share from which the data was backed up, Phoenix may log an "Access is denied" error in the Phoenix restore logs. This issue can occur for both SMB and NFS shares due to any of the following reasons:

  • If the proxy fails to back up the access control list (ACL).
  • If the proxy backs up ACL of all the users but access permissions change in the interval between the backup and restore.

Phoenix successfully restores the share data and users can access the files and folders irrespective of the difference in access permissions.




The forward slashes (/) in the NFS share name appear as underscores (_) under the Progress Logs tab of the backup jobs.
For example, the NFS share name /vol/ISO appears as vol_ISO, instead of /vol/ISO, on the Progress Logs tab.






Phoenix does not back up shares and folders with names containing Unicode characters.  




The backup and restore job count stats are not displayed in the Administrative group details for NAS shares. The stats calculation will be available in the subsequent releases.



MS-SQL Server Issues

Issue Description

Scheduled differential backups in scenarios where databases are added, deleted, or renamed, are treated as full backups. However, the Backup Status Report displays the type of backup as "Differential Backup". 

Note: To know the scenarios in which Phoenix treats a differential backup as a full backup, see Differential backups


If you cancel an ongoing differential backup, then the subsequent restore to the original instance will fail.


Try restore to an alternate location.

Try full backup after you cancel a differential backup


The database backup fails if it contains small IDF and MDF files (<1MB) and the auto growth for the database file is set to a small value. IDF and MDF files (<1MB) and the auto growth for the database file is set to a small value.


Change the auto growth for the databases to a reasonably large value (~100MB).


When Phoenix service is restarted after database and metadata upload but VSS records incomplete backup, all subsequent differential backup jobs fail until the next full backup is successful.


The restore operation of database files fails on a File server. 

Workaround: Ensure that you restore the database files on the MS-SQL server.


A scheduled backup may create a job and leave it in the queued state if the Phoenix agent installed on the preferred secondary node gets disconnected and its state is not updated to Phoenix.


The Restore to point in time page displays an empty popup when you click the Calendar icon to select the date to restore databases or leave the database in the standby mode.


Ensure that you set the date manually in the box to restore databases or leave the database in the standby mode.


If a location other than C:\Program Files\Druva was selected at the time installation of the Phoenix agent, auto-upgrade from the client versions older than <current-2Apr-client-version> installs Phoenix at the default install location, that is,  C:\Program Files\Druva and not at the custom installation location.


After restore to FCI, if a database has the "read-only" property set, the read-only property value is not preserved in the restored database.


The substring search for the availability group on the Jobs page does not display any job records.


The snapshot restore from the SQL Server 2012 to SQL Server 2016 in the Standby mode fails with the SQL7 error.

ROBO-10837 Transaction mark restore can fail with SQL8 error if the transaction marks were recorded before a full backup. 


The Druva Phoenix Resource Status Report displays incorrect operating system and client version details for the SQL availability group.

VMware Issues

Issue Description


If Backup Proxy loses connectivity with Phoenix Cloud or vCenter Server, the connection status under the Backup Proxies tab appears as "Connected".


Wait for some time to allow the status to change. 


If you changed the name of the Backup Proxy virtual machine using vSphere Client, the changed name does not reflect under the Backup Proxies tab (Manage VMware Infrastructure > <ESXi or ESXi or vCenter Server>). 

Note: This is by design. Backup Proxy continues to work as expected even if the name change does not reflect on Phoenix Master Management Console. 


Sometimes, when you restore a virtual machine, the virtual machine properties reflect the default values. 


The restored virtual machine fails to boot with error "Operating system not found". 
After the restore, the boot sequence changes as VMware gives priority to IDE. VMware gives priority to IDE. 


After the restore, change the boot sequence manually.


If you register an instance of Backup Proxy for an ESXi or ESXi or vCenter Server using its IP address, and you register another instance of Backup Proxy for the same ESXi or ESXi or vCenter Server using its FQDN, the ESXi or ESXi or vCenter Server appears twice in the Manage VMware Infrastructure list, once with the IP address, and once with the FQDN. 


A backup that starts at the same time as the snapshot creation of a virtual machine from the vSphere Client fails with the following error:

Connection dropped before completing backup.error


Start the backup after the snapshot creation completes. 


If you attempt to delete a snapshot of a virtual machine from the vSphere Client at the time when backup of that virtual machine is complete, the delete operation does not delete the snapshot that Phoenix previously created.

Note: This is by design. Concurrent operations (in this scenario, deleting the snapshot that Phoenix created, and deleted the snapshot that was created from the vSphere Client) on the same virtual machine are not supported. 

ROBO-1130 If you added vService Dependency to a virtual machine, backups and restores of this virtual machine complete successfully. However, the restored virtual machine does not contain the vService Dependency. 


If backup proxy loses connectivity when the backup of associated virtual machines is in progress, the backup proxy attempts to establish connection multiple times and its status under the Backup Proxies tab appears as "Pending". 


Cancel the pending backup task and then wait for next scheduled backup, that will be successful.


Restores of virtual machines having one or more  Unicode characters in their names is not supported. 


If a virtual machine has many disks, and you attempt to restore this virtual machine, some UI elements on the Restore Data window might not appear as expected. 


Backups from virtual machines associated with a backup proxy fail if the VMware tools reset the time of the virtual machines to a value that is likely incorrect even if the time synchronization check box is not selected. 


Disable time synchronization. For a complete set of instructions, see Disabling time synchronization in the VMware library. 

ROBO- 1481

Backup fails with an unknown error " #100011001".


Retry backup.


Backup fails with the error "device_key" for a virtual machine that has 60 SCSI disks.

ROBO -1531

Backup fails if in vRDM disks the same LUN is mapped on two different disks.


Unmap the LUN from both disks and then map to only one of the disks.


Before restore, if the VRDM disk is changed to an independent VRDM, the guest repair will fail.


VMware restore fails when you try to restore a virtual machine/Disk Restore from Higher Version VC to lower version VC.

ROBO- 1572

Restore fails when the bus sharing mode of a virtual machine is modified. The backup will also fail if the bus sharing mode is not set to "None".


The backup proxy utility command does not work after upgrading backup proxy.


  1. Find the currently installed RPM.
    Command: rpm -qa | grep druva

  2. Remove existing RPM.
    Command: rpm -e <curr_rpm_name>

  3. Install the new RPM.
    Command: rpm -ivh <new_rpm_name.rpm>


Virtual machine restore fails with the error "vim.fault.FileAlreadyExists". 
In case of standalone ESX, if the restored virtual machine is deleted from the inventory of its folder and VMX is in the backend.


Rename the folder in the backend in order to restore the virtual machine. 


While configuring the proxy server with the agent or CloudCache, if you misspell the proxy type, the proxy will still be configured successfully with default Socks5 protocol.


During backup, if you update the proxy configuration then the ongoing backup will fail.


If the Phoenix service restarts multiple times during FLR, the initiated restore may fail if the restored folder contains files that cannot be deleted.


While activating the Backup Proxy, if it does not get connected to the VCenter. The Backup proxy will remain in the not connected state till we restart the Backup Proxy service.


The Backup activity report, displays the first VM  backup as Incremental instead of full.


If you delete a Backup Proxy during a VM backup, the detailed log request will fail with an Internal error.


If a rule (for example Rule1) is applied to a folder (for example Folder1) in a data store, and a rule (for example Rule1a) is applied to its subfolder (for example Folder1a).

  • In the Rule1a, already configured VMs are excluded from auto-configuration (Exclude already configured VMs setting is checked.)
  • In the Rule1, already configured VMs are included for autoconfiguration. (Exclude already configured VMs setting is unchecked.)

If there is a virtual machine (for example VM1) in Folder1a that was manually configured. Rule1a does not auto configure VM1, but VM1 becomes eligible for auto-configuration using Rule1. However, VM1 is not configured using Rule1. 


A full restore to an original location of an encrypted virtual machine does not overwrite a VMDK file.


Delete the detached VMDK file after restoring the encrypted virtual machine.


In Linux, if the root folder ("/") is selected for a file restore, the file level restore job ends with the message "Successful with error" and the files are not restored. 


The backup proxy OVA fails to deploy on a standalone ESXi version 6.5 using the Web UI. This is a VMware issue. Druva is working with VMware to resolve this issue. 




Phoenix does not support backup and restore for virtual machines with disks using NVMe controller. 

Note: Backup and restore for virtual machines with disks using NVMe controller will be supported in a future release. 

ROBO-7164 If a backup proxy version 4.6.xx or earlier is upgraded to a backup proxy version 4.7 or later while a restore to an alternate location is in progress:
  • The restore stops with an incomplete restore
  • The incomplete restore job does not resume after the update is complete. The restore job restarts after the update is complete and the backup proxy creates another virtual machine at destination hypervisor
  • The backup proxy ends up creating two virtual machines at the destination hypervisor instead of resuming the original incomplete restore


Phoenix generates empty audit trails for bulk operation failures.


Virtual machine backup and restore fails if the sum of NBD and NBDSSL connection buffers to an ESXi host exceeds 32 MB. 

If backup or restore fails due to insufficient memory, it is possible that buffer size of the NBD and NBDSSL connections to the ESXi host is more than 32 MB. 

To fix this issue, add the following entry in the Phoenix configuration file on the Phoenix backup proxy virtual machine: 

Path to the Phoenix configuration file on the Phoenix backup proxy virtual machine: etc/Phoenix/VMWARE/Phoenix.cfg

ROBO-12018 Default file exclusion fails for Windows Server 2003 virtual machines. Files excluded from backup of Windows virtual machines are: pagefile.sys and hiberfil.sys.
PHN-4716 (SQL Server aware virtual machine backups) Phoenix is unable to back up a database, if the database has a blank space before or after its name. 

DRaaS Issues

Issue Description


If you restart a DR proxy service while the DR restore is in progress, the DR restore status will reset to 0%.


Cloud DR for LVM is only supported if the LVM has a file system.


If you have disabled the DR status of the virtual machine and then update the server group, then the DR status for that virtual machine will automatically be enabled.


The DR restore job will fail if cache network is disconnected.


If you have scheduled an “immediate after backup” option for the DR policy, and after a successful backup the proxy is down, then the job will not be created for DR.


The AWS account number on the Manage DR policy page does not match with the AWS account number on the Update DR policy page.


If you configure a virtual machine in the region where DR proxy is absent, then in such scenario, the virtual machine Update Server Group page does not display any policy name in the DR policy tab.


If you have backed up a virtual machine with distributed vSwitch and you are restoring that virtual machine on ESX that does have access to distributed vSwitch then the restore fails with an Internal error "VMWARE_VMOMI65535".


For scheduled backup job retry, if backup job switches on multiple Backup Proxies, then the requests for job logs generate multiple but same Log Request jobs.


Phoenix DRaaS fails to create an instance for a queued DR failover job for a virtual machine. The instance is created once the job is in the running state.


Phoenix DRaaS fails to clean up the DR instance when the network connection breaks during the initial part of the failover.


Phoenix fails to mark the status of the DR failover job for a virtual machine as stale or failed after 24 hours.


An ongoing failover job fails if you upgrade a Phoenix AWS proxy with version 4.8.0 or 4.8.1 to 4.8.2.


If you restart the Phoenix client service after launching failback, Phoenix marks the failback job as failed. However, if a target virtual machine exists, the failback agent does not delete the failed back virtual machine.


After failback, Phoenix shows an incorrect guest operating system of the target virtual machine on the VM summary page.

  • After you delete a DR plan, it is displayed in the list of DR plans, but the entity name is linked. 
  • After you rename a DR plan, the new name is not reflected on the Jobs page for the older jobs. 
PHN-10727 If you remove a VM from a DR plan while a DR restore job is running, the job is marked as canceled. 

CloudCache Issues

Issue Description

While editing the CloudCache details on the Edit CloudCache Details page, if you select the Backup to Cloud if unable to backup to Cache option and click Save, the checkbox still appears clear. However, the backup will continue to cloud if CloudCache is not connected.

You can view the current state (Enabled/Disabled) of the Backup to Cloud if unable to backup to Cache option on the Summary tab of the Manage CloudCache page.


Group administrators do not receive an email notification if the CloudCache attached to the group is decommissioned.


After the license expires, Phoenix allows you to add volumes to the existing CloudCache. However, Phoenix disallows server backups.


Phoenix does not generate an audit trail for the updated retention period of a backup set attached to a CloudCache.

Hyper-V issues

Issue Description


Backups fails for virtual machines if Backup integration service (volume checkpoint) is unavailable for the guest operating system.


If the length of the virtual machine files path exceeds 256 characters, VSS snapshot creation fails that results in a failed backup job.


The backup and restore job count stats are not displayed in the Administrative group details for Hyper-V. The stats calculation will be available in the subsequent releases.



ROBO-7746 Backup fails with HYPERV28 for a virtual machine using both clustered and non clustered storage. This error occurs due to VSS issues.


Phoenix generates empty audit trails for bulk operation failures.

ROBO-8919 If the FLR proxy fails to remove the folder it earlier created at the UNC share location, both file-level restore and retry fail. If Phoenix service is restarted while the FLR job is running, Phoenix removes the folder created on UNC share and retries running the restore job. If somehow Phoenix fails to remove the folder it created earlier, the file-level restore skips all the files.
ROBO-8898 Hostname of the FLR proxy deployed on Windows Server 2008 host is not updated on Phoenix management console after a user changes it.
ROBO-9041 Virtual machine restore to an alternate location fails if the user entered virtual machine name contains any of these special characters: (/ \ : " ? * < > | "
ROBO-9884 Ensure that the English language pack is installed on the Hyper-V hosts with a certified operating system that is localized to a language other than English.  If the English language pack is not installed, the Phoenix Management Console is unable to list the number of virtual machines running on the host. 

For more information, see System requirements. To add the English language pack on Windows, see Available Language Packs for Windows. [External link to Microsoft documentation]


Restore fails for virtual machines:

  • Running on Windows Server 2008 Hyper-V hosts
  • That use the SCSI controller for the virtual disks 


To resolve this issue:

  1. Download the virtual disks.
  2. Manually create a virtual machine on the Windows Server 2008 Hyper-V host. 
  3. Attach the downloaded virtual disks to the virtual machine.
ROBO-11225 If a backup job is running on a client and an administrator runs an auto-upgrade, the running backup and the client upgrade jobs freeze. 
  • Was this article helpful?