NetVault Plug-in for VMware 11.1

In Job Name, specify a name for the job. Assign a descriptive name that allows you to easily identify the job for monitoring its progress or restoring data.

A job name can contain alphanumeric and non-alphanumeric characters, but it cannot contain non-Latin characters. There is no length restriction. However, a maximum of 40 characters is recommended on all platforms.

In the Selections list, select an existing Backup Selection Set, or complete the following steps to create a set:

Click Create New to open the NetVault Backup Selections page.

Open the NetVault Backup Client on which the plug-in is installed, and then open VMware Plugin.

Open the applicable ESX/ESXi or vCenter Server. Depending on your VMware setup and inventory view type, open the applicable container nodes.

NOTE: To switch between the two views (Hosts and Clusters and VMs and Templates), select the ESX/ESXi or vCenter Server, and in the Actions list, click Toggle Inventory View. This option is only available when server node is open.

Select the data that you want to back up:

•

Back up all virtual machines within a container: Select the container node. Data selection is possible at all levels, starting from the Datacenter node. For example, to back up all virtual machines hosted on an ESX or ESXi server, select the host node; in the VMs and Templates view, select the corresponding folder node.

•

Back up individual virtual machines: Open the applicable container nodes (for example, Datacenter, Cluster, Resource Pool, and other nodes), and select the virtual machines that you want to back up. In the VMs and Templates view, open the folder node, and select the virtual machines that you want to back up.

You can also select the container node, and then clear the check marks for the virtual machines that you do not want to back up.

•

Back up individual virtual disks: Open the virtual machine node, and select the disks that you want to back up. The virtual disks are named “Hard Disk 1,” “Hard Disk 2,” … “Hard Disk n.” The plug-in only lists those disks for which it can generate a snapshot.

You can also select the container node (for example, Datacenter, Resource Pool, ESX Server, or folder) or the individual virtual machines, and specify the disk type (system or data) in the Backup Options Set. For more information about this setting, see Virtual machine disk selection options.

The virtual machine configuration files, “.nvram”, and log files are always backed up whether you select the entire virtual machine or its individual disks.

Click Save, and in the Create New Set dialog box, type a name for the set.

A set name can contain alphanumeric and non-alphanumeric characters, but it cannot contain non-Latin characters. On Linux OS, the names can have a maximum of 200 characters. On Windows OS, there is no length restriction. However, a maximum of 40 characters is recommended on all platforms.

Click Save to close the dialog box.

In the Plugin Options list, select an existing Backup Options Set, or complete the following steps to create a set:

IMPORTANT: On Windows, when you create a Backup Options Set for image‑level backups, you must use the default set “Default Backup Options — VMware Plugin — Windows” as the template. If you use any other set as the template, the backup may fail.

Click Create New to open the VMware Plugin Backup Options page.

Under Backup Type, select one of the following options.

Table 1. Image-level backup types for virtual machines
Option	Description
Full	Select this option to back up all allocated sectors on a virtual disk.
Incremental	Select this option to back up disk sectors that have changed since the last Full, Differential, or Incremental Backup.
Differential	Select this option to back up disk sectors that have changed since the last Full Backup.

For more information about image-level backup types, see Image-level backups.

IMPORTANT: After reverting to a snapshot that is older than its last Incremental Backup, you must perform a Full Backup of that virtual machine before performing Incremental Backups again. If you perform an Incremental Backup, the job reports a filefault error and fails. For more information, see http://kb.vmware.com/kb/1021607.

Under VM disk selection options, select one of the following options.

The disk selection option is applied to all virtual machines that you have selected explicitly by selecting the corresponding nodes or implicitly by selecting the container node (for example, ESX Server, Resource Pool, Datacenter, or folder). This option is not applied to virtual machines for which you have explicitly selected one or more virtual disks in the selections tree.

Table 2. Virtual machine disk selection options
Option	Description
Back up all disks	To back up all available virtual disks for the selected virtual machines, use this option. NOTE: The plug-in only backs up those disks for which it can generate a snapshot.
Exclude boot disk	To back up only the data disks and exclude the boot disks for the selected virtual machines, use this option.
Exclude data disks	To back up only the boot disk and exclude the data disks for the selected virtual machines, use this option. NOTE: To identify the boot disks, the plug-in only considers the active partitions on the MBR disk. Moreover, the plug-in only considers the first disk that meets the active boot partition criteria as the boot disk. The plug-in does not handle machines with multiple operating systems.

Under Other Options, configure the following settings.

Table 3. Other backup options for Plug‑in for VMware

Option

Description

Enable Changed Block Tracking for VM

By default, CBT is disabled for a virtual machine. If you want to perform Incremental or Differential Backups of virtual machines, you must enable CBT for the Full Backup that serves as the base backup for subsequent Incremental and Differential Backups.

To enable CBT on all virtual machines included in the backup job, select this check box. (You can also enable CBT on specific virtual machines by using the Enable Change Block Tracking method. For more information, see Enabling CBT on individual virtual machines.)

If you select this check box and the plug-in fails to change this setting on a virtual machine, a warning message is logged.

If you do not select this check box, the plug-in does not change the CBT setting on the virtual machines during backups. Depending on whether CBT is enabled or disabled, the plug-in uses the appropriate backup method (CBT-based Full, Incremental, or Differential or non-CBT Full Backup) to back up the virtual machines.

Note the following:

•

To create the first Full Backup of a virtual machine with Changed Block Tracking enabled, the target virtual machine must be powered off to enable CBT. For more information, see http://kb.vmware.com/kb/1031873.

The virtual machine can be restarted as soon as the backup job starts transferring data. During subsequent Full, Incremental, or Differential Backups, the virtual machine can remain powered on.

•

At the time of enabling CBT, no snapshots should exist on the target virtual machine. For more information, see http://kb.vmware.com/kb/1033816.

•

Once you enable CBT for a virtual machine, all subsequent Full, Incremental, and Differential Backups are performed as CBT-based backups for that virtual machine. You do not have to select this option for each job.

To stop using CBT, you must disable this feature. For more information, see Disabling CBT on a virtual machine or Manually disabling CBT on a virtual machine.

Perform File Level Indexing

When you perform an image-level backup, the plug-in automatically generates file-level indexes for all supported volumes that are included in the backup snapshot. File-level indexing allows you to restore individual files and directories from Full, Incremental, and Differential image-level backups of virtual machines.

File-level indexing is available to volumes that use the following file systems:

•

Windows: NTFS

•

Linux and UNIX: EXT2, EXT3, and EXT4

The Plug‑in for VMware also supports volumes managed by Logical Volume Manager (LVM) on Linux-based systems and Logical Disk Manager (LDM) on Windows-based systems as single or spanned disks.

NOTE: The current version of the plug-in does not support Windows Server 2012 ReFS (Resilient File System) and Striped Disks.

File-level indexing does not affect the backup size. However, it increases the backup index size and the total backup time. The amount of time taken to perform file-level indexing depends on several factors, including the number of files, fragmentation of files on the volumes, network traffic, and load on the ESX or vCenter Server.

If you do not want to use image-level backups for file-level restores, clear this check box.

Enable Active Block Mapping

The Active Block Mapping (ABM) technology provides filters to remove unused blocks during backups. Removing the unused blocks reduces the backup size and the amount of data transferred over the network. You can use ABM with CBT to back up only active and changed blocks during Incremental and Differential Backups.

ABM is supported on the NTFS file system on basic disks and EXT file systems.

ABM is disabled by default. Selecting this check box enables ABM, which allows the plug-in to back up only those sections of the virtual disks that are active. ABM scans the disk and detects inactive blocks; these blocks are then omitted during backups.

NOTE: ABM also removes deleted data. When you restore a virtual machine that was backed up using ABM, you cannot perform undelete operations for that virtual machine. Without Active Block Mapping, CBT includes the deleted blocks.

If a backup job includes any unsupported disk types, this option is ignored for those disks.

Remove snapshots from previous backups

When you run a backup job, the plug-in creates a snapshot named “BKB_SNAP” on the virtual machine to back up the selected data. Regardless of whether the backup completes successfully or fails, the plug‑in removes the snapshot on job completion. The plug-in also removes the snapshot if the job is aborted or the parent or child processes are terminated for any reason. However, if a job exits abnormally, the cleanup process may fail to remove the snapshot. To address such cases, the plug‑in provides an option to delete the snapshot when you run the job next time.

You can select the Remove snapshots from previous backups check box to remove any existing snapshots when you run the current job. Only the snapshots named “BKB_SNAP” are removed from the virtual machines. This option does not remove any other snapshots that exist on the virtual machine.

Maximum Number of Parallel Streams

By default, the plug-in generates a single data stream for a backup job, and backs up the selected virtual machines in a sequential manner. To increase throughput and reduce overall backup time for image-level backups, you can configure the plug-in to perform parallel backups of multiple virtual machines.

This setting determines the maximum number of parallel data streams that can be generated for an image-level backup job. For example, if 10 virtual machines are included in a job, and you set this parameter to four, the plug-in tries to back up four virtual machines in parallel.

The actual number of parallel streams for a job depends on the following factors:

•

The number of backup devices or streams available to the backup job. For example, if you set this parameter to 4 and only two tape drives are available or a NetVault SmartDisk can only support two concurrent streams, the plug-in only processes two virtual machines concurrently.

•

The number of virtual machines included in the job. For example, if you set this parameter to 4, and the number of virtual machines selected for backup is 3, the plug-in creates only three data streams.

For parallel backups, the plug-in generates a parent process that coordinates the overall backup and individual child processes that perform the actual task of backing up the virtual machines. The maximum child processes that can be generated for a backup job is equal to the Maximum Number of Parallel Streams configured for the job. The parent and the child processes are all created on the NetVault Backup Client on which the plug-in is running.

When a child process acquires a backup device and creates a backup stream, it is assigned a task to back up a virtual machine. After the task completes, the process is assigned the next task if there are any more virtual machines to be backed up. Each task is assigned a Task ID. A snapshot is generated only when a virtual machine is assigned to a child process.

When configuring Maximum Number of Parallel Streams, consider the following:

•

Verify that the value configured for this option does not exceed the number of backup devices or streams accessible to a job.

If the available devices are insufficient, more than one child process may try to access and write to the same device, causing significant increase in the overall backup time.

•

The load of running multiple processes on a NetVault Backup Client can have a negative impact on performance.

•

The load of backing up multiple virtual machines from the same datastore can increase the I/O activity on the datatsore.

•

The load of backing up multiple virtual machine that are managed by the same ESX Server can increase the load on the host if you are using the LAN Transport mode (nbd/nbdssl) for data transfer.

NOTE: Only a single child process is used to back up the multiple disks of a virtual machine even if additional data streams are available. The child process creates a snapshot of the whole virtual machine and retains the snapshot until the job completes regardless of whether the whole virtual machine or its individual disks are backed up.

NOTE: This setting is not applied to backup jobs created with plug-in versions earlier than 2.5. For such jobs, the plug-in generates a single stream and backs up the virtual machines sequentially.

Enable Restartable backups

This option allows you to restart a job that completes with some failed virtual machines. The restarted instance backs up only those virtual machines that failed previously; virtual machines that are backed up successfully are not included in the restarted instance.

When a restartable backup completes with some failed virtual machines, the plug-in generates a backup index for the completed virtual machines and sets the job status to Job Stopped. The log messages and log context show which virtual machines have failed for the job. When you restart the job later, the plug-in runs an Incremental Backup job to back up the failed virtual machines.

You can restart a job from the Job Status page. For more information about this method, see Restarting a backup job. A restarted instance does not back up virtual machines that are added to the host after the initial job is run.

NOTE: If all virtual machines fail for a job, the job status is set to Backup Failed. You cannot restart a failed backup job.

Click Save, and in the Create New Set dialog box, type a name for the set.

Click Save to close the dialog box.

Select or create the Schedule Set, Target Set, and Advanced Options Set. For more information about these sets, see the Dell NetVault Backup Administrator’s Guide.

NOTE: When performing multi-stream backups, do not select the Ensure backup is the first on the target media check box. If you select this check box for multi-stream backups, each data stream targets a separate piece of media to exist as the first backup on the media item. Thus, if a backup generates five streams, the job tries to obtain five blank or new media items.

This option does not apply to disk-based storage devices.

To submit the job for scheduling, click Save & Submit. You can monitor the job progress from the Job Status page and view the logs from the View Logs page.

To save the job definition without scheduling it, click Save. You can view, edit, or run this job from the Manage Job Definitions page. This job is not displayed on the Job Status page until you submit it.

For more information about Job Status, View Logs, and Manage Job Definitions, see the Dell NetVault Backup Administrator’s Guide.

Additional notes

•

When you select a container node (for example, Datacenter, Resource Pool, or ESX/ESXi Server), any new virtual machine that is added to the host after a job has been defined is automatically included in the backup. Similarly, when you remove a virtual machine from the host, it is automatically excluded from the backup. You do not have to modify the job definition each time a virtual machine is added to or removed from the host.

•

If a selected virtual machine is moved to a different host using Storage vMotion after a backup job has been defined, it is backed up as part of that job only if the following requirements are met:

•

A vCenter Server has been added to the plug-in, and the access to the virtual machine is being authenticated through the vCenter Server.

•

The ESX Server Host to which the selected virtual machine has been moved to is also controlled by the same vCenter Server.

If these requirements are not met, the plug-in fails to back up the virtual machine that has been moved.

•

If Fault Tolerance (FT) is enabled for a virtual machine, the plug-in does not allow you to select it for a backup.

If this feature is enabled after including a virtual machine in a backup job, that virtual machine is not backed up and the message “Cannot back up VM — Fault Tolerance is enabled for the VM” is added to the job logs.

•

You can select virtual machines that are part of a Virtual Application (vApp) along with the other virtual machines. Currently, the selection tree does not provide any indicators to distinguish between such virtual machines from other virtual machines that are not part of a vApp.

If you select a vApp node, only the virtual machines contained in the vApp are backed up. The metadata for the vApp is not included in the backup.

•

If you add a new virtual machine to the Backup Selection Set (or to an ESX Server that is included in the Backup Selection Set) after the initial Full Backup or any subsequent Incremental or Differential Backup, the plug-in does the following:

•

Creates a Full Backup of the new virtual machine if you run Incremental or Differential Backup.

•

Creates a non-CBT Full Backup, unless CBT is already enabled for the virtual machine or you have selected the Enable Change Block Tracking for VM check box for the Incremental or Differential Backup.

•

Runs subsequent Incremental Backups of the new virtual machine as Incremental if CBT is enabled and a CBT-based Full Backup has been performed for the virtual machine.

•

Runs subsequent Differential Backups of the new virtual machine as Full Backups because these backups are based on the last Full Backup created for that Backup Selection Set.

Example:

Select any one virtual machine (for example, MyVM1) and create a Backup Selection Set (for example, MySelectionSet).

Perform a CBT-enabled Full Backup using MySelectionSet.

Modify MySelectionSet and add a virtual machine (for example, MyVM2).

Perform CBT-enabled Incremental and Differential Backup using MySelectionSet.

Perform another CBT-enabled Incremental and Differential Backup using MySelectionSet.

Result:

•

When you select Incremental Backup, the plug-in creates a Full Backup of MyVM2 in step 4 and an Incremental Backup of that virtual machine in step 5.

•

When you select Differential Backup, the plug-in creates a Full Backup of MyVM2 in step 4 and step 5 because these backups are based on the last Full Backup created in step 2.

•

When using CBT for backups, virtual machine migration using Storage vMotion or VMware vSphere® Storage DRS™ is not recommended. For more information, see http://kb.vmware.com/kb/2048201.

•

If you perform a Full Backup of a Thick Provisioned Lazy Zeroed disk without enabling CBT, the empty sectors are converted to actual zeros during backup. When you restore the backup, the disk type is changed to Eager Zeroed.

•

After backing up the data, the plug-in writes the backup index to the NetVault Database. If an error occurs during this step (for example, if there is a file transfer error or the server runs out of disk space), the plug-in reports the following error:

Log message: Failed to write index of backup to the database.

Log context: The index for this backup may be retrieved and added to the database by scanning the backup media.

In such cases, you can scan the backup media to recover the index. However, before importing the index, you must review the log messages to verify that no other errors were reported during the backup.

Backup selection tree icons

Table 4. Backup selection tree icons
Icon	Description
	vCenter Server
	Datacenter Server
	ESX or ESXi Server Cluster
	Closed folder
	Open folder
	ESX or ESXi Server
	ESX or ESXi Server in maintenance mode
	Inaccessible ESX or ESXi Server
	Virtual Appliance (vApp)
	Resource pool
	Virtual machine (powered on)
	Inaccessible virtual machine
	Mounted virtual machine
	Suspended virtual machine
	Powered off virtual machine
	Virtual machine (powered on, CBT enabled)
	Inaccessible virtual machine (CBT enabled)
	Mounted virtual machine (CBT enabled)
	Suspended virtual machine (CBT enabled)
	Powered off virtual machine (CBT enabled)
	Primary virtual machine (powered on) in a Fault Tolerant group
	Inaccessible primary virtual machine
	Suspended primary virtual machine
	Powered off primary virtual machine
	Secondary virtual machine in a Fault Tolerant group
	Inaccessible secondary virtual machine
	Suspended secondary virtual machine
	Powered off secondary virtual machine

Restarting a backup job

The Plug‑in for VMware provides the ability to restart a job to back up only those virtual machines that failed previously. To use this method, you must configure the backup option Enable Restartable backups for the job. For more information about this option, see Performing image-level backups. When a restartable backup completes with some failed virtual machines, the plug-in generates a backup index for the completed virtual machines and sets the job status to Job Stopped. You can restart this job from the Job Status page.