Currently Being Moderated

VMware vSphere Data Protection 5.1 Deployment

Posted by kuver Singh on Mar 5, 2013 8:46:02 AM
VMware DATA Recovery:
Virtual machines are important resources that change on a daily basis. You should create backups for all
virtual machines – or VMs – in your environment. The information normally backed up for VMs includes
the operating system, applications, and data. All of this information for each VM resides in one or more
virtual machine disk files, which have the vmdk extension. In addition to virtual machine disk files, a
VMware virtual machine has a configuration file, which has the vmx extension. These configuration files
contain information about the VM configuration such as how many CPUs the VM has, how much
memory the VM can use, and which vmdk files belong to the VM. As with physical machines, virtual
machines need to be backed up periodically to recover from loss of data due to human error, technical
malfunction, etc.
INTRODUCTION:-vSphere Data Protection (VDP) is VMware’s backup and recovery solution designed for small- to
medium-sized organizations.
VDP offers several key features to make virtual machine backup and recovery rapid, reliable,
manageable, and affordable. It provides a simple interface to quickly create, configure, and schedule
backup jobs. It is an agent-less, pure disk-based solution, allowing for faster restores over solutions that
write directly to tape.
A VM that utilizes an agent for backup and recovery requires the VM to be powered on to perform a
restore using that agent. With VDP, this is not the case - backups and recoveries can be performed
regardless of the VM’s power state.
VDP provides a centralized management interface through VMware vCenter Server so scheduling of
backup and recovery jobs are performed using the VMware vSphere Web Client. The vSphere Web
Client allows an administrator to create, configure, schedule and execute backup and restore jobs
through an intuitive wizard.
Additionally, VDP enables the recovery of an entire virtual machine image for virtual machines running
any operating system. It also enables the recovery of individual files and directories for virtual machines
running Microsoft Windows or Linux operating systems. Multiple restore points for each virtual machine
are displayed to easily select a specific restore point.
VDP utilizes built-in data deduplication technology to help minimize disk space consumption.
Deduplication eliminates duplicate storage blocks as backup data written to disk. Therefore, vSphere
administrators can maintain multiple point-in-time copies of VMs using only a fraction of the storage
that would normally be required versus storing multiple full copies of a VM.
The built-in data deduplication works in conjunction with Changed Block Tracking in vSphere to
minimize the amount of data to process. On vSphere hosts running 4.0 and later, a record of each virtual
machine’s changed blocks is maintained. This is called Changed Block Tracking. VDP utilizes this
information and backs up only the blocks that have changed since the last backup of the VM. This helps
reduce the amount of data to be transferred from the virtual machines to the VDP appliance for backup.
COMPONENTS:-Now, you will look at components that are part of VDP solution and understand their functions.
VDP is a virtual machine appliance deployed from an .ova or open virtual appliance file.
The VMware vStorage API for Data Protection (VADP) is utilized by VDP. This includes the Changed Block
Tracking (CBT) mechanism. CBT tracks the changes made to a VM at the block level and provides this
information to VDP so that only changed blocks are backed up. This significantly reduces storage
consumption and speeds up backup and recovery times with VDP.
VMware Tools on Windows contains Microsoft Volume Shadow Copy Service (VSS) components to assist
with guest OS and application quiescing when backing up Windows VMs.
The VDP appliances have three backup data destination storage capacities: .5 TB, 1 TB, 2 TB.  Note that
the actual amount of storage capacity consumed by the appliance will likely be greater than the backup
data destination storage capacity. Please see the VDP Administration Guide for more information.
VDP is managed using the vSphere Web Client. Please note that the traditional vSphere Client cannot be
used to manage VDP – you must use the vSphere Web Client.
The VDP appliance is deployed by default with four vCPUs and 4 GB of RAM.
For the capacities of 0.5, one, or two terabytes, the amount of disk space (thick-provisioned) actually
consumed by each appliance is 850 GB, 1.6TB and 3.1TB respectively. Thin provisioning can be used, but
if it is, the administrator should actively monitor disk consumption.
It is important to note that once the VDP appliance is deployed, additional storage cannot be added.
The VDP appliance guest OS is SuSE Linux 11.
vCenter Server 5.1 or higher is required to use VDP.
VDP can backup VMs running on vSphere hosts 4.0 and higher. The VDP appliance connects to the hosts
using port 902. If there is a firewall between the VDP appliance and the vSphere host, then port 902
must be opened.
VDP management must done via the vSphere Web Client.
Before installing the VDP appliance, it is recommended that you review the VMware Data Protection
Administration Guide for a complete list of prerequisites.
CONSIDERATION:-When establishing VDP in your environment, there are special considerations to be aware of. Here are a
few of them:
VDP does not support tape integration. It is a disk-based backup and recovery tool. It utilizes disk-based
storage to enable speedy recovery of your virtual machines.
VDP cannot backup virtual machine protected by VMware Fault Tolerance – or FT. This is because there
is no way to snapshot FT-protected virtual machines.
Each instance of vCenter Server can support a maximum of 10 VDP backup appliances per vCenter. Each
VDP appliance supports backing up a maximum of 8 VMs concurrently.
VMware supports protecting a maximum of up to 100 VMs per appliance.
VDP appliances do not share information with each other. For example, if two VDP appliances are
configured to protect the same VM, this could cause errors if both appliances try to back up the VM at
the same time.
The VMware Data Protection Administration Guide lists VDP prerequisites in detail, but you will take a
brief look at them now.The VDP appliance is installed on a vSphere host version 4.x or 5.x.The host that
runs the backup appliance must be managed by vCenter Server 5.1.The virtual machines to be backed
up and the backup appliance must both connect to a vSphere host using port 902. If there is a firewall
between the appliance and the vSphere host, port 902 must be open.The VMware Client Integration
Plug-in needs to be installed in your browser. Before installing VDP, it is recommended that you review
the VMware Data Protection Administration Guide for a complete list of prerequisites.
PRE-INSTALLATION Steps:-Before deploying VDP, it is important to properly size the appliance, as additional storage cannot be
added after deployment. Sizing requirements are based on the number and types of virtual machines,
the amount of data and retention periods, and data change rates. You can find general sizing guidelines
in the Administration Guide.
DNS should be properly configured for vCenter Server and all vSphere hosts. A DNS host record for each
VDP appliance should be added prior to deployment. Also, NTP must be configured for all vSphere hosts
and vCenter Server.
A user account with administrative permissions in vCenter and the SSO server is required to deploy
VDP.With all these ready, you can proceed to deploy the VDP ova file.
Installation Procedure:
1 Select vCenter Home > vCenter > VMs and Templates. Expand the vCenter tree and select the vSphere
Data Protection appliance. Right-click the appliance and select Power On.
2 Right-click the appliance and select Open Console.
3 After the installation files load, the Welcome screen for the vSphere Data Protection menu appears. Open
a web browser and type:
https://<ip address of VDP appliance>:8543/vdp-configure/
4 From the VMware Login screen, enter the following:
a User: root
b Password: changeme
c Click Login
5 The Welcome screen appears. Click Next.
6 The Network settings dialog box appears. Specify (or confirm) the following:
a IPv4 Static address
b Netmask
c Gateway
d Primary DNS
e Secondary DNS
f Host name
g Domain
7 Click Next.
8 The Time Zone dialog box appears. Select the appropriate time zone and click Next.
9 The vSphere Data Protection credentials dialog box appears. For vSphere Data Protection credentials type
in the appliance password. This will be the universal configuration password. Specify a password that
contains the following:
? Nine characters
? At least one uppercase letter
? At least one lowercase letter
? At least one number
? No special characters
10 Click Next.
11 The vCenter registration dialog box appears. Specify the following:
a vCenter user name (If the user belongs to a domain account then it should be entered in the format
b vCenter password
c vCenter host name (IP address or FQDN)
d vCenter port
e SSO host name (IP address or FQDN)
VDP 5.1 Procedure:-
1 Log in to the vSphere Web Client and select vCenter > Datacenters.
2 On the Objects tab, click Actions > Deploy OVF Template.
3 Select the source where the vSphere Data Protection appliance is located.
4 By default the select source dialog is set to OVF Packages. Change it to OVA Packages.
5 Select the appliance and click Open.
6 After the appliance .ova file is selected, click Next.
7 Review the template details and click Next.
8 On the Accept EULAs screen, read the license agreement, click Accept, and then click Next.
9 On the Select name and folder screen, enter the name for the appliance and click on folder or datacenter
you want it deployed in. Click Next.
10 Select the host for the appliance and click Next.
11 Select the virtual disk format (“Impact of Selecting Thin or Thick Provisioned Disks” on page 47 provides
additional information) and the location of the storage for the appliance. Click Next.
12 Select the Destination Network for the appliance and click Next.
13 In the Customize template, specify the Default Gateway, DNS, Network 1 IP Address, and Network 1
Netmask. Confirm that the IP addresses are correct. Setting incorrect IP addresses in this dialog box will
require the .ova to be redeployed. Click Next.
NOTE:- The vSphere Data Protection appliance does not support DHCP. The appliance requires a static IP
14 On the Ready to complete screen, confirm that all of the deployment options are correct and click Finish.
vCenter deploys the vSphere Data Protection appliance. Monitor Recent Tasks to determine when the
deployment is complete.
DATA Deduplication and Changed Block Tracking
Data deduplication works in conjunction with the Changed Block Tracking functionality on vSphere hosts
to minimize the amount of data to process. The blocks that have changed since the last backup are
tracked, so that at the time of backup only those blocks are presented for backup. VDP leverages this
functionality reducing the amount of data to check and transfer.
Note that Changed Block Tracking is only available for virtual machines running with hardware version 7
or later. If you are backing up virtual machines created with older versions of VM hardware, they will
take longer to back up because they do not use Changed Block Tracking during backup.
Changed Block Tracking comes into play only after the initial full backup of a virtual machine has been
performed. The second time a backup is performed, VDP backs up only those blocks that have changed
since the last backup. This helps reduce the amount of space consumed by full backups and the storage
I/O is also reduced as only the changed blocks are backed up.
If all the virtual machines that are being backed up are running the same version of the operating
system, this will maximize the efficiency of the de-duplication algorithm. Deduplication is evaluated for
all virtual machines protected by the same VDP appliance. To maximize the benefit of de-duplication,
you should back up similar virtual machines using the same VDP appliance.
How Vsphere Protects DATA
This diagram shows what happens when you back up a virtual machine with vSphere Data Protection
First, the VDP application creates a quiesced snapshot of the virtual machine. Different versions of
Windows may use different quiescing methods such as Microsoft's Volume Shadow Copy Service (VSS)
or the File Sync Driver that comes with VMware Tools.
Second, VDP opens the deduplication store and backs up the virtual machine files. When the backup
starts, VDP checks to see if the virtual disk has been previously backed up to the VDP appliance. If it has
been backed up, then VDP uses the Changed Block Tracking information to backup only the blocks that
have changed since the last backup. If the Changed Block Tracking information has become invalid, then
a full backup is performed.
Third, the incremental changes are applied to the previous backup to create a synthetic full backup in
the de-duplication store.
Network-Based Copy Versus SCSI HOT-Add
In the previous screen, you saw that vSphere Data Protection (VDP) can perform a backup utilizing SCSI
HotAdd (pronounced skuzzy hot add) or a network-based backup. SCSI HotAdd is the process of
attaching the virtual disk file being backed up directly to the VDP appliance. This helps reduce the
amount of time required for the backup and keeps backup traffic off of the network.
VDP tries to use the SCSI hot-add method first. If that method fails, then VDP automatically falls back to
a network-based copy.
There are several conditions under which SCSI hot-add is known to fail:
The first is if SCSI HotAdd feature is not licensed on the vSphere host.
Hot-Add will fail if the datastore hosting one or more of the virtual machine’s disks is not accessible to
the vSphere host running the backup appliance.
Another reason Hot-Add will fail is if there are no free virtual SCSI slots left on the backup appliance. The
backup appliance supports up to 4 SCSI controllers. Each controller can support 15 SCSI devices. This
allows for 60 SCSI disks being attached. The backup appliance uses one of those slots for its boot disk
and one for each VMDK deduplication store.
Restoring VMs and File-Level Restore (FLR)
vSphere Data Protection creates backups of entire virtual machines. These backups can be restored in
their entirety using the vSphere Web Client for vSphere Data Protection.
However, if you only want to restore specific files from these virtual machines, then you use the vSphere
Data Protection Restore Client.
The Restore Client allows you to mount specific virtual machine backups as file systems and then browse
the file system to find the files you want to restore. The Restore Client operates in one of two modes:
Basic—allows you to only mount backups that were made from the machine you are logging in with, and
any files that you restore will be restored to this client. For example, if you were logging in to the
Restore Client in Basic mode from a Windows host named “WS44” then you would only be able to
mount and browse backups of “WS44.”
Advanced—allows you to mount and browse any backups that are contained in vSphere Data
To restore a virtual machine, the first step is to select the source from which to restore the failed virtual
machine. You can choose to restore a single virtual machine or more than one virtual machine. You can
specify collections of virtual machines, such as all virtual machines in a datacenter, or select individual
virtual machines. You can use filters to view only selected virtual machines and then restore them. If
multiple restore points are selected for a single virtual machine, VDP restores the virtual machine to the
most recent restore point selected. Note that a restore point is a point in time for which a backup was
VDP also features File Level Restore for Windows and Linux guests as long as the volumes backed up
were one of the following formats: Windows NTFS and Linux LVM, Ext 2, Ext 3 supported - basic disks
The VDP Restore Client makes it convenient and easy for end-users to perform self-service restores.
Using the URL shown on the screen, a user can log in (administrative permissions required) and browse
the restore points available. All restore points can be viewed or you can filter by date.

kuver Singh

kuver Singh

Member since: Jan 17, 2011

The real definition of the phrase cloud computing simply means the ability to access files and applications online through multiple devices—your computer, browser, or phone. VMware/Microsoft has already been offering cloud services through the web...

View kuver Singh's profile


Recent Comments

No recent comments.

Create Your Own Personal Blog

To create a personal blog on, sign into your account, click on "Manage Account" in the top right corner of any page, click on the "Blog Posts" tab and then click on "Create a Personal Blog" or "Write a Blog Post" from within your account profile.

Note: All blogs will be monitored and reviewed for content. Any blogs not related to virtualization or considered to be spam or offensive will be removed.