First off I wanted to say Welcome and Thank you to PHD Virtual, one of the newest sponsors of It's companies like the ones to the right of this post helping to keep the community content flowing by putting out awesome products within the VMware ecosystem.  That being said with the addition of PHD Virtual as a blog sponsor I decided to write a sponsored review of there newest backup software PHD Virtual Backup v5.4.

PHD Virtual is not a new company to me.  I actually evaluated their backup solutions back when it was called esXpress in the 3.x days.  Let me first say that PHD has come a long way with their backup product since then.  Back in those days there were packages that were actually installed on the ESX hosts and it was all managed through SSH.  Now, with the release of PHD Virtual Backup 5.4 everything is fully integrated into vCenter and much much more easier to use.   Anyways, enough about my history with PHD and on to the review….

My Environment

I first wanted to mention that all of this was tested on a laptop with 8 GB of RAM running VMware Workstation and some cheap, slow NFS storage :).   Obviously things would perform a bit nicer in a production environment but this is what I have to work with for the time being.  Also, PHD supports an architecture that utilizes multiple appliances in order to improve performance and scalability, however my tests were done utilizing a single appliance.  Certainly go and get yourself a copy of PHD Virtual Backup and try it in your own environment to see its' full potential.


Honestly this step probably doesn't even need a walkthrough since it is so simple, but in the interest of being thorough I'll talk about it.    

First off you need to install the Backup Console and the vCenter plugin onto the machine that you wish to manage PHD from (normally your workstation). This is essentially three clicks in a Next Next Done fashion and your off to the races. 

Next you need to deploy the virtual appliance (the brains of the operation).  Again, this is very easy.  Simply go to 'File->Deploy OVF Template' and select your PHD Virtual Backup appliance.  I accepted all defaults as far as resources go and was up and running in less than a minute.  Once the appliance is up and running then only step that needs to be configured at the VM's console is the networking which is done by hitting CTRL+N on the console.

That's it!  You have a copy of PHD Virtual Backup installed now.  A few notables on the appliance

  • Appliance runs Linux and Postgres – To me this is as good as gold.  No windows licenses to take care of, no MSSQL lisense to worry about.
  • The fact that its an OVF – It's nice to just import the appliance and configure, rather than have to install a base OS, patch the base OS, install all the pre-requisites for the software, then pull down the software and install it.  It's all done in one simple step for you.


So now that we have everything installed it's time to configure the appliance by attaching some storage to it, setting up email settings, retention settings etc..  Here is where the vCenter integration shines.  Right-click any object in vCenter and select PHD Virtual Backup -> Console.    After the console has loaded all of the configuration is done by clicking on (you guessed it!) configuration.  Having that integration right inside of vCenter is very nice, for instance you can kick off a backup or replication job by accessing PHD's items on the right-click context menu. 

For the most part configuration can be broken down into the same categories as the tabs that it displays; General, Backup Storage, Network, Email, Backup Retention, Replication, and Connectors.

The 'General Tab' isn't very exciting.  It's basically used as you can see from the screenshot to setup some of the more basic settings such as your time zone, NTP settings as well as vCenter credentials.

The 'Backup Storage Tab' is used to setup your actual storage that you are going to use for your backups.  You can attach backup storage to an appliance through either a directly attached vmdk, CIFS, or NFS.  For my testing I chose to use NFS.  Also in this section you can set some percentages to trigger warnings or actually stop backing up depending on free space available as well as enable/disable compression on your backups.

The 'Network Tab' is pretty self explanatory.  Allows you to change the appliance's IP address or edit/configure DNS settings.

The 'Email Tab' is just that.  Allows you to setup your SMTP server as well as sender and receiver information.

The 'Backup Retention Tab' is where you define how many and how long you want to keep your backups.  There are a couple of predefined configurations (Keep all and Typical), but you are also able to customize your retention based on how many recent backups to keep and in addition to that how many daily, weekly, monthly and yearly backups to maintain.

The 'Replication Tab' is where you will setup your pointers to your appliances backup storage.  Meaning you specify which backups you would like to replicate to another appliance.  Replication is explained in more detail further down in this post.

The 'Connectors Tab' is where you would setup the shares in order to use the Backup Data Connector (BDC).  The BDC gives you the functionality to access your backups in their native format in order to archive to a tape or disk.  I didn't get a chance to fully test this but I like having the ability to do so.

So for the most part that is all the configuration that you need to do.  So far so good, a very easy install and quite an intuitive configuration.  Now it's time to get into the thick of the application and do some backups and restores.


Getting a backup job created and running is a very simple task in PHD Virtual Backup.  You can either set the backup job from within the console or (again with the vCenter integration) right click the VM you want to backup and select PHD Virtual Backup -> Backup.

 If choosing the first option you have a few ways to backup your VMs.  Selecting the Hosts & Clusters view will allow you to select a datacenter, cluster, host or specific VM.  The VMs and Templates view allows you to select a Datacenter or a folder from within your vCenter inventory.  

Next you will need to select which PHD appliance you would like to back the VMs up to.  Since I'm testing this out on a workstation lab with only one appliance this is a pretty easy decision for me, but if you were in production you would have the ability to deploy multiple appliances attached to different backup storage in order to distribute your backups across different types or tiers of storage.

From there you need to schedule your backup (Now, Once, Daily, Weekly) and move on to the backup options.  Here you have the ability to be run a verify task (on all blocks or just new blocks) on your backup job.  You can also check whether to include powered off VMs, mark the backup as archived (meaning it will not be deleted by the retention policy), chose whether to use VMware's change block tracking and whether or not to quiesce the OS if backing up and Windows guest.

I ran a few different jobs on various OSes (Linux and Windows).  The backup speed was great, the compression and deduplication were great.  Obviously that first initial seed takes a bit of time to complete, but from then thereafter if utilizing CBT the subsequent incremental backups are very efficient and sometimes occur in less than 5 minutes depending on the number of changed blocks. Also worth mentioning, for each job you are able to get a nice detailed view as to what happened behind the scenes in regards to the speed and deduplication.   Again, a few notables

  • Once again, very easy, very intuitive – documentation isn't needed in order to start using this PHD's software.  The step by step wizards guide you through everything.  If you are familiar with vCenter and vCenters inventory views you should be fine using this application.
  • Retention options are great – I like the ability to archive or save weekly\y/monthly/yearly backups. 
  • Retention is applied to the VBA – Here's a wish list item.  It would be nice to apply different retention settings to individual VMs within a job or to the job itself. 


Restoring is handled in two different fashions by PHD Virtual Backup.  You can restore the complete VM or perform a file-level restore to just get individual files.  The full VM restore is pretty straightforward and once again easy to perform.  You simply need to find your restore point you wish to restore inside of the backup catalog and click 'Restore' along the top toolbar.

The Restore Job wizard doesn't require much from a user perspective.  Simply select your storage or where you want to place the restored VM, and select a few options in regards to the networking on the newly restored VM (which network to attach it to and whether to generate a new MAC address).  Just as the backup process my testing with full VM restores went flawless.  

The file-level restore is handled by setting up an iSCSI target to the backed up VM on the PHD appliance, then connecting your client to that iSCSI target using the Windows iSCSI Initiator. You can kick of a FLR wizard in a few ways, either by selecting your VM within the Backup Catalog or clicking 'Create' in the File Recover section of the application.  Either way will invoke the wizard.  There are really only a few steps within the wizard…


First, select your VM and restore point that you wish to grab the files from.  It looks simple in this screenshot but if you have 100's of VMs with 100's of restore points there are some great filters at the top that allow you to narrow down your search.  

Secondly you can chose to mount the created iSCSI target to your local machine by simply checking the box.  If you are going to be restoring the files on your local machine immediately this is the best option, otherwise if you need to mount the target later you can do so by clicking 'Mount' on the File Recovery section.

Once the wizard is complete, if restoring from a Windows guest you should be able to see the folders and files within explorer.  Linux guests require a separate third party application (ext2explore) to be downloaded and installed in order to perform FLR


  • Full VM restore – works flawless and very easy to perform.
  • FLR for Linux VMs – you need to download a separate application in order to perform and FLR on your Linux guests.  Would love to see this fully integrated into the application.
  • Exchange item level recovery is available via a partnership with Lucid8.  If you have time check out this video which takes you through it.


Technically I don't have the gear to fully test replication so I will just explain how replication works within PHD Virtual Backup.  Unlike most products that perform their replication on the VM level, PHD actually performs their replication on the backup level, meaning that your production VMs are not affected at all.  Essentially once a VM has been backed up, as long as you have added your backup storage as a replication target, you can now perform a replication on the backup files of that VM.   Replication is meant to occur from site to site, or from appliance to appliance, giving you a true DR solution in the event of a site failure.


  • Only changed blocks are replicated, making the replication just as efficient as the backup process.
  • Built in test mode allows you to completely test the functionality of the replicated VM in a standby environment.  This is great, just as the Verify Backups option is on the backup jobs.  What good is a backup or a replica if it doesn't work?  This way, you will know it does.


Certainly if you are looking for a backup solution for your vSphere environment check out PHD's Virtual Backup.  The software is very easy to use and is fully integrated into vCenter making it very easy to access what you need when you need it.  Installation is a breeze, configuration is simple and quick.   Utilizing CBT, the incremental backups run extremely fast which will allow you to minimize backup windows.  I like the different approach to replication which leaves production VMs untouched and affects only the backups.  Also, the scalability of this product is great and when using multiple appliances throughout the network you can really distribute the load as much as you want.   As with any product there are certainly some spots where that they could improve on or features that they could add such as support for more granular control per VM inside of a backup job and the ability to add more than one type of backup storage to an appliance.  That being said however, PHD Virtual Backup does a tremendous job at doing what it was built for; backing up your virtual infrastructure.  A very intuitive interface makes this application a breeze to install, set up, configure and to use.  I can guarantee that you won't need a training course to figure this one out 🙂  Certainly don't just rely upon me, go grab a trial version from PHD and check it out yourself.