NetVault 13.1.1 - Release Notes

On the Manage Devices page, in Tree View, clicking Scan starts scanning the selected slot, drive, or library immediately. It does not show the Minimum life for imported backups (days) dialog box or confirmation message.

NVBU-14745 (32197)

On the Manage Devices page, in Tree View, you cannot select multiple media from the list of available media for a bulk blank operation.

NVBU-14758 (32214)

On the Manage Devices page, in Tree View, you cannot select multiple media from the list of available media for a bulk label operation.

NVBU-14759 (32215)

When performing backups with the NetVault Plug-in for Teradata, if the number of parallel streams exceeds the SoftStreamLimit set for the target NetVault SmartDisk, the backup job stops responding.

NVBU-389 (1948)

When a backup that is configured as “restartable” is duplicated, only the data backed up in the last restarted instance of the job is duplicated.

NVBU-1160 (3971, (NVG‑4022)

In NetVault with a non-English locale, the Running Status field does not display a current job status as the output of nvreport command and Report Job.

NVBU-1186 (3997) (NVG‑4206)

On Linux- and UNIX-based systems, the progress message that is displayed during NetVault installation is displayed correctly only if the system is configured to use the UTF-8 character set.

NVBU-10588 (26249)

If the Network Manager process (nvnmgr) stops responding due to any error and fails to exit properly, it prevents the NetVault Service from starting on that machine.

Workaround: If this error occurs, manually stop or end the nvnmgr process before restarting the NetVault Service.

NVBU-11005 (26778)

After upgrading to Firefox 31, if you try to log in to the NetVault WebUI, the following error message is displayed:

Issuer certificate is invalid. (Error code: sec_error_ca_cert_invalid)

The new security library security.use_mozillapkix_verification raises this certification error when you try to access a self-signed certificate.

Workaround: To resolve this issue, do the following:

Type about: config in the Firefox address bar.

Locate the preference security.use_mozillapkix_verification, and set it to false.

To save the setting, close the browser window.

NVBU-11321 (27189)

On Windows, if a copy of a “.cfg” file is created in the <NetVault home>/config directory, the WebUI may display empty or partly empty dialog boxes when the settings are accessed from the Change Settings page after restarting the service.

NVBU-12734 (29149)

Under certain conditions, the NetVault Server may be unable to determine the status of the push installation task on a remote client. In such cases, the task status may remain in the “Running” state. This error can occur under the following conditions:

•

The deployment agent (nvclientinstaller) exits unexpectedly.

•

The deployment agent fails to update its completion status to the server due to unknown errors.

•

The deployment agent fails to send the status update messages to the correct IP address. This error can occur if the NetVault Server is running on a multihomed machine and the preferred network address is not configured for the server.

Subsequent tasks remain in a “Pending” state until you cancel the previous task.

NVBU-13610 (30508)

When performing push installations, if you use any illegal characters in the machine password for a new client, the remote installer reports an error and fails to install the software on the client.

NVBU-13632 (30537)

If the server IP address is used to configure a local package store (for example, \\10.11.12.3\PkgStore), the user credentials are not validated when the package store is added or updated. If you specify invalid credentials, no errors are reported. However, the deployment task fails if the authentication fails and the packages cannot be retrieved from the store. Therefore, Quest recommends that you use the server name when configuring the path to a local store (for example, \\WinServer1\PkgStore).

NVBU-13647 (30558)

Known issues related to Plug-in for RDA on Solaris:

•

Backing up file system data from Solaris clients to DR Series Systems: When you perform file system backups to DR Series systems, the backups run in the Passthrough mode by default. You can change the default setting on the DR Series system to run backups in the Dedupe mode, but you may experience slow backup performance in this mode.

•

Restoring data from a DR Series System to a Solaris Client: When you run multiple restore jobs at the same time, it can affect the performance of individual jobs and result in low aggregate throughput.

NVBU-13678 (30596)

•

When performing optimized replication between two DR Series systems, if the source DR Series system becomes unavailable, a normal data copy operation using the NetVault Server or Client is performed when the device becomes available again.

•

When performing a backup using the NetVault Plug-in for VMware (Plug-in for VMware), if you set the client mode type to dedupe, the throughput is lower when compared with the passthrough mode. If you do not set the client mode on the DR Series system, the Plug-in for RDA automatically chooses the best mode for the backup.

•

Restore to a NFS location fails when a Duplicate saveset created using the optimized replication option is used for the job.

On a Linux system, NetVault fails to access a DR Series system if you do the following:

Install the Plug-in for RDA.

Install NetVault.

Stop NetVault Service.

Remove the Plug-in for RDA.

When you remove the plug-in, it deletes the directory /usr/local/oca-libs, which causes an error when NetVault tries to access the device.

Workaround: After removing the plug-in, complete the following steps:

Ensure that the /usr/local/oca-libs directory has been removed. If not, remove the directory.

To create a symbolic link, type the following command:

ln -sf /usr/netvault/dynlib/oca-libs /usr/local/oca-libs

Restart the NetVault Service.

The Delete Non-Scheduled Jobs option is intended to delete unscheduled jobs after a specified period. Because the jobs using the Triggered schedule type have no physical scheduled time, they are also deleted after the specified period elapses. If you are using Triggered jobs in your NetVault environment, do not use the Delete Non-Scheduled Jobs option.

When using NetVault Plug-in for NDMP (Plug-in for NDMP), if you run multiple jobs with the auto-deletion option set to “On,” the parent and child jobs may still be listed on the Job Status page.

Workaround: To correct this issue, close the NetVault WebUI, and start it again. There is no need to stop and restart the NetVault Service.

The canned report “Disk Storage Devices — General” is intended for the NetVault SmartDisk devices; this report does not show details about the DR Series systems and Data Domain Systems.

After restoring the NetVault Database, when you restart the NetVault Service, the job status shows “Scheduler Aborted Whilst Job Running.” This message can be ignored.

On a Solaris (SPARC or x86-64) system, when the LANG variable is set to en_US.UTF-8, the restore job wizard does not display the filenames on the Create Selection Set page. This issue is not encountered when LANG is set to C.

If the default locale is changed (for example, from UTF8 to EUC or from EUC to UTF8), the EULA may not be displayed properly.

Workaround: To view the EULA, use a terminal that can display EUC character set.

Beginning with NetVault Backup 11.1, if Plug-in for Oracle or Plug-in for Informix is installed on a Linux machine where NetVault Server is also installed, the backups of the plug-ins may fail due to certain libraries loading.

Workaround: Add the NetVault library path to the NetVault startup.sh script. For example, on a Linux-based NetVault Server:

LD_LIBRARY_PATH=/usr/netvault/lib:/lib:/usr/lib:$LD_LIBRARY_PATH

export LD_LIBRARY_PATH

NVBU-14424

(31721)

Pre-configured reports, including Disk Storage Devices - General, Front Side Data, Storage Utilization, and Storage Utilization - RAS, did not report the correct value for Licensed Protected Capacity.

NVBU-14623

Job details are not displayed when you click a policy job in the Job Status widget on the NetVault Dashboard page.

NVBU-17538

After running a job, when you go to the Job Status page to monitor the status of the job, it shows all the jobs with previously applied filters instead of showing all current activity.

Workaround: Select the Current Activity option in the View By job status filter option.

NVBU-17582

You cannot restore the data items of a network share backup located using the catalog search feature.

NVBU-17632

If you perform multi-stream backups to tape, select the Archive and Migration options in Backup Life, and identify an expiration date, the savesets are corrupted after the expiration is reached.

Workarounds:

•

When you scan the media to recover data, select the Migrate (Discard) Original option.

•

Take the index offline and perform the restore.

NVBU-18297

When a LUN is mounted on an unsupported volume and the disk information is retrieved for a snapshot, the disk properties shows for VMware disk and not Huawei or Compellent disk. This causes the snapshot to fail.

NV-47, NV-49, (NVBU-18642, DNFSP-1016)

In the NetVault WebUI, when editing the Notification Profile, Extra Environment variables do not display.

NVBU-18925

In specific instances, changing the time window in the server monitor from 8 hours to another time range then back to 8 hours results in an incorrect graph.

Workaround: Use the 12-hour or 24-hour setting.

NVBU-18975

The Unusable Media widget on the dashboard does not show media that must be recovered.

NVBU-19214

The error “duplicate key value violates unique constraint” message appears multiple times in the pg_sql log files.

NVBU-19246

For LTO-7 tapes, the Explore Storage page inaccurately displays the size of an unused tape.

NVBU-19250

When running a historic job report by date on a Japanese Standard Time server, not all information displays in the report.

NVBU-19313

The Storage Utilization report shows the wrong licensed capacity for SmartDisk.

NVBU-19385

While choosing a default plug-in option after switching the language, the Web UI goes into deadlock, and performs no further operations.

NVBU-19427

When you try to reset the locale from French to English from the Server Change Setting page, the NetVault Server goes into Safe Mode.

NVBU-19493

On Linux, when creating a Large_VTL, if the WebUI times out due to inactivity or the user closes the WebUI while the creation is still in progress, the VTL creation does not occur.

Workaround: In the WebUI, create a VTL using the “Re-add previously generated virtual device” option, and then, in the webservice.cfg file, increase the SessionExpiryTime.

NVBU-19625

When a large number of jobs are in Running state, the Job Status page hangs.

NVBU-19694

When using file immutability, removing a full backup saveset with no immutable files also removes persistent immutable Incremental Backup savesets.

Workaround: In the Advanced Options > Backup Life section, do NOT select Make Backup Immutable. If you select this option for an incremental backup but you do not select it for a full backup, if you try to delete the saveset, then an error occurs.

NV-2044, NV-2045

When used with NetVault Core 13.1.x, plug-ins that use workflows in SCHEDULE_ARGS format may crash. Specific plug-ins include Plug-in for Oracle, Plug-in for SAP on Oracle, Plug-in for SAP Hana, Plug-in for NDMP, and Plug-in for PostgreSQL.

NV-3230

Table 4. QoreStor known issues
Known issue	Issue ID
NetVault job fails if it uses the same container with data stubbed for new jobs.	NV-1428
An Plug-in for Oracle restore job passes silently with errors when it should fail because the data is still on a QoreStor glacier tier.	NV-1430
A Plug-in for NDMP restore job succeeds when it should fail immediately as the data is archived with QoreStor.	NV-1431
QoreStor does not support multi-tenancy.	QS-437, DNR3‑71
If you run parallel backup or restore jobs on an AIX client with a QoreStor target, the jobs fail with a “Job Manger has lost the message channel unexpected” error message.	QS-4252, DNR3-870

Table 5. Command-Line Interface known issues

Known issue

Issue ID

When the Broadcast details to machines on local networks setting is disabled on a client, the nvclient utility fails to add the client.

Workaround: Use the nvclientadd -clientip <Client IP address> command to add the client.

NVBU-12966 (28829)

The asf_load_media command fails if the machine name contains any spaces.

When the asf_release_drive command is issued to various brands of library devices (for example, those containing various Exabyte or 9840 drives), a delay of up to eight (8) minutes may occur before a piece of media is unloaded.

Workaround: The workaround is to issue the “Unload” command from the NetVault WebUI.

In the Navigation pane, click Manage Devices, and then in the list of devices, click the applicable drive or the corresponding Manage Device icon.

On the Tape Drive Management page, click Unload.

The drive remains in and “Idle” state, and media are not unloaded as expected. If the logs for the selected device are viewed, an entry stating “Unload media in drive <X> failed — Device unavailable” is displayed. Repeat Step 2 to unload the media.

The range function does not work properly for the nvacslsmedia command.

Workaround: The workaround is to explicitly run this command for each media item that you want to allocate or deallocate. You can use a script to do this task.

When a space character is all that is specified for a library name with the nvcleandrive command, the first library on the Manage Devices page performs a clean operation on the specified drive. For example, if you run the command nvcleandrive libraryname “<space character only>” -librarydrivenumber <X>, NetVault automatically targets the first library on the Manage Devices page and perform a clean operation on drive <X>.

When a null value is specified with the nvexpiresaveset command, any random saveset is expired.

When you run the nvjobmodify command, the job is submitted to run, even if you do not include the “‑submit” option in the syntax.

When a null string is specified with the nvremovemedia command, the first entry is deleted from the Media Database.

The nvsetcleaninglives command does not report any errors if you configure incorrect values for the lives option.

The “-include” option for nvsetcreate cannot be used to specify data selections below the “Fixed Drives” node in the selection tree.

The nvsetmodify command cannot be used to modify Schedule Sets. You can create a set using nvsetcreate, and assign it the same name to overwrite the existing set.

On Windows-based installations, the nvsyncronizesilomedia command does not update the media status on the Manage Devices page.

In an environment where the NetVault Client is installed on a Linux machine and the NetVault Server is installed on a Windows machine, if a non-default user password contains eight or more characters, the Command-Line Interface (CLI) utilities fail.

Workaround: Reset the user password to one that contains seven or fewer characters.

NVBU-17861

When used to select nodes for backup and restore, the nvsetcreate command does not interpret a single quote and instead changes it to a blank space.

Workaround: To avoid this issue, write the string argument into a parameter file and use the nvsetcreate -parameters switch to specify the file.

NV-2157

Table 6. Plug-in for FileSystem known issues

Known issue

Issue ID

On Snap Appliance Filers, data must only be relocated to existing directories. Items relocated to directories that do not exist cannot be accessed due to permissions settings on the newly created directories.

DNFSP-113 (13979, (FSG‑188)

On ZFS, if the backup data contains ACLs on extended attributes, the restore job fails when you use the default restore options. If errors occur on restore of metadata, select the Discard Metadata option to restore the data.

DNFSP-456 (14324, (FSG‑514)

On Solaris ZFS and UFS file systems, the files and directories for which only the extended attributes have changed do not qualify for Incremental Backups.

DNFSP-636 (14504, FSG‑684)

When restoring a Linux or UNIX backup to a Windows-based NetVault Client, the items are restored successfully, but the job generates the following warning and log context:

Log message: <num> items and their contents could not be restored

Log context: <filename>: writing to file: Incorrect function

These messages can be ignored.

DNFSP-718 (14586, FSG‑760)

NetVault does not allow Apple Xsan restores when there are no affinity tags on the target server.

DNFSP-727 (14596, FSG‑770)

The exclusion list is not enforced for a corrupt file in an Incremental Backup.

DNFSP-910

On Oracle Solaris hosts, enabling a multi-stream FileSystem backup causes Incremental Backup performance to appear similar or worse than a Full Backup.

DNFSP-921

If the name of a network share folder includes Japanese characters, NetVault does not add it.

DNFSP-932

The FileSystem backup and restore job fails to write an entry into the log file when the Path to Backup log and Path to Restore log options are selected.

DNFSP-991

On Windows 2008 and Windows 2003, the System State node does not include the Removable Storage Manager Database node.

The Verify After Backup option does not work for a restarted job instance.

When you submit a policy job created for a UNIX file system on a Windows system, the job status shows “Completed with Warnings,” but no data is backed up.

The following known issues apply to the Backup Windows File Short Names and Restore Windows File Short Names options:

•

Restoring short names only works if an existing file with the same short name does not exist.

•

Attempts to back up or restore short names fail if you enable Windows-based Compression or Encryption option on a target system.

Although VSS backups complete successfully, the message “Error deleting volume snapshot(s): VSS_E_BAD_STATE” is written to the NetVault logs.

If a job restores a Windows symbolic link (file created using mklink, not Windows shortcut), where the symbolic link file exists on the target file system and the data file the symbolic link targets also exists, the data in the targeted data file is lost. Ensure that the data files targeted by symbolic links are included in backups. Quest recommends that you delete the existing symbolic links that you want to restore from the target file system before running a restore job. To avoid any risks of data loss, you can also perform backup of data files possibly at risk before running the restore job.

When a symbolic link is restored, GuardianOS applies the Xattr attribute user.SnACL to it. This behavior is expected and does not have any negative effects.

Backups on GFS and GFS2 can stop responding if the files or directories are frozen.

In an environment with many direct automounts on Solaris Clients, a backup job may remain at “Waiting for Job Options” status while the Plug-in for FileSystem attempts to load the automounts temporarily to determine whether they are on remote systems. This delay can be alleviated by selecting the “Backup through Remote Mounts” option. If necessary, clear or exclude remote mounts or add the following stanza to the file nvfs.cfg:

[Performance:DoRecursiveAutoMountScan]

Value=FALSE

Restore operation of multi-volume selection for an offhost backup fails after successful off-host backup for multi-volume selection.

NV-48 (NVBU‑18349)

Table 7. Managed Service Provider (MSP) known issues
Known issue	Issue ID
The nvreport utility does not work in the MSP environment.	NVBU-18346
A tenant user can see and delete other tenant users’ Plug-in Options and Source sets.	NVBU-19795