This is important because today we have released Veeam Backup & Replication v6 after over a year of R&D effort. What makes this release so significant in my view is that it’s your release. When I say it’s your release I mean that if you take a look at all the new features and enhancements, they were almost all driven by direct customer and partner feedback.

Better Scalability? You got it!
Our new proxy/repository architecture allows you to scale out a single Backup & Replication server to thousands of VMs. You can still go with a single server and let it be the proxy and repository but as your environment grows you can now grow with it.

Enhanced Replication? Yep!
If you take a look at the What’s New around replication alone you’ll see that this is one of the areas we focused most of our attention on. Re-IP? Check. Failback? Check. Traffic Throttling? Check. Better seeding? Check.

Support for multiple hypervisors? Yes!
Now I know not everyone asked us to support Windows Server Hyper-V but a fair number of customers and partners did, as well as a lot of Hyper-V shops that wanted the functionality of Veeam Backup & Replication. The fact that you get support for VMware and Microsoft virtualization all from a single install and interface is huge news.

I'd like to give a big Thank You to all of our customers, partners and prospects who had so much to do with this release. We get our best feedback through our forums so please continue to post great feedback there. We're already working on the next release and we'll have some BIG news to share early next year (again, from YOUR input) so watch for the countdown clock.

For a comprehensive overview of v6, please check out the webinar series that includes product demos: http://vee.am/v6webinars

We released our hotfix on 20 October, 2011, that means it took us a full 57 days to release a hotfix for Veeam Backup & Replication to FULLY support vSphere 5. One of the main issues with Veeam Backup & Replication 5.0.2 was that it didn’t work for all processing modes with VMSF5 (just like every other vendor at the time of this writing). A processing mode popular with our customers is direct SAN access utilizing the vStorage API for Data Protection (VADP). In order to support direct SAN access for VMFS 5, we had to move to the latest version of the VDDK API which was only made available the same day as the GA release of vSphere 5.

57 days to incorporate a new version of a low-level API
1368 hours to fully test
82,080 minutes to fix things that worked previously

So, rather than trying to shout from the rooftops that we supported vSphere 5 the day it was released (or even weeks before), we wanted to make sure everything worked. We make backup and DR software after all, and we don’t want our customers getting into a mess with their production data based on claims by marketing. I think Anton did a good job of explaining our position in the forums, please take a moment to read his post.

Before you go checking our downloads page for the hotfix, please be advised that it’s only available through Veeam support. Since this isn’t a full release, and not all of our customers will need it, we are only making it available via support. Of course, Veeam Backup & Replication v6 will have full support of vSphere 5 when it’s released later this year, and that will be our official release vehicle for vSphere 5.

We have repeatedly asked Tolly who is sponsoring the report but they have declined to tell us. As of this writing, Veeam has no idea who is paying Tolly for this report (but we have our suspicions). So, in an effort to be truly fair and accurate, we proposed the following to Tolly:

• Veeam co-sponsors the tests and the report, and pays half the sponsorship fee;
• Veeam participates in determining the test criteria on an equal basis with Tolly and the other sponsor; or, alternatively, test criteria are determined by a mutually agreed-upon neutral third party;
• Both vendors’ representatives are present during the tests;
• The tests are conducted in September 2011. Veeam’s key people who can participate in the project are on vacation or on other projects until September 15; and
• The tests are conducted with Veeam Backup & Replication v6, which will be released in Q4 2011.

The above proposal was not accepted by Tolly, and we were told the process was too far along with the sponsoring vendor. They did tell us that we could sponsor our own report at a later time if we like.

My question now is how fair and accurate will this report be? According to the Fair Testing Charter (parts below, emphasis mine), we will be allowed to review the findings and provide feedback.

Finally, we provide them [Veeam] with the opportunity to give us an official commentary on their results, the essence of which is included in any documents we might publish.

It’s interesting they say the “essence of which.” To me it means that they will listen to our feedback but will edit it into an “essence” that puts their client in a better light. Likewise, “might publish” says to me that if Veeam were to come out ahead in the test, even with cherry-picked criteria, the report will never see the light of day.

It appears that Tolly has been doing this type of “fair and accurate” testing for quite some time; I just hope no one loses millions of dollars this time based on Tolly’s findings: http://www.brisbanetimes.com.au/business/compressed-into-nothingness-20100808-11qdh.html?from=age_ft (the Tolly report is available here)

Steve Duplessie from ESG has a great take on head-to-head testing and why ESG doesn’t do it: http://www.thebiggertruth.com/2010/03/the-politics-of-testing/

And I particularly like Simon Bramfitt’s assessment of Tolly: http://blog.simonbramfitt.com/2011/06/tolly-trumps/

Vendors in the backup space have also had their share of problems with Tolly, including Quest: http://www.infoworld.com/d/data-management/just-when-you-thought-backup-wars-were-dead-654?page=0,0[BB1]

It’s clear to us that one of our competitors intends to use this Tolly report as part of a competitive marketing campaign against us. Veeam did NOT participate in the testing and, from our vantage point, the criteria have been designed to avoid testing typical customer scenarios as well as very features that make Veeam Backup & Replication v5 stand out as the best solution on the market. So please treat any report from Tolly as suspect.

Once you configure the iSCSI initiator, you will see your production VMFS LUNs appear in the disk management snap-in of the Veeam Backup & Replication server.  Be sure to check out this is this post from Justin’s IT Blog about this topic. This is shown in the figure below:

This would require any zoning on the iSCSI target to include the iSCSI qualified name (IQN) of the Veeam Backup & Replication server to the storage provisioned for the vSphere (or VI3) environment. In this example, the storage is a 2 TB LUN that is formatted with VMFS. While the drive path is visible within Windows, do not try to initialize or format these LUNs within the disk management snap-in, as this could corrupt or overwrite data stored on the VMFS LUNs. Further, note that the Veeam Backup & Replication v5 and later setup automatically disables the automount feature of Windows. Automount allows for automatically mounting and assigning configuration to newly connected volumes. If you add a VMFS datastore to the Veeam Backup & Replication server, with automount enabled, the operating system may initialize and re-signature the volume. This would make it unrecognizable by the ESX(i) hosts.

Having these LUNs visible within disk management can ensure that all of the required LUNs are available to Veeam Backup & Replication, including viewing the target and LUN IDs as presented from the storage processor. Conversely, if all VMFS LUNs are not visible; there may be a zoning issue.

Direct SAN access processing mode allows Veeam Backup & Replication to communicate directly with the storage for the highest backup job performance. Further, if the backup target supports iSCSI or fibre channel; direct SAN access mode also enables a completely LAN-free backup implementation.

In the case of iSCSI storage, this usually means that the virtual machine that Veeam Backup & Replication is running on will have a presence of at least two (or more) networks (vNICs) as in most iSCSI storage networks are separated from other networks (including production networks), at least at a VLAN level. Should fibre channel storage be used as the backup target, NPIV can be leveraged to connect a virtual machine running Veeam Backup & Replication directly to the SAN fabric.

What configuration practices have you done with Veeam Backup & Replication as a virtual machine and iSCSI storage networks? Share your comments below.

These three options are very important in how the backup deduplication is implemented. First of all, this deduplication algorithm is based on the source. This means that the deduplication algorithm is applied before any transfers are made from the ESX(i) host to the backup or replica of a virtual machine. To implement this, a backup or replication job has three choices in the job definition:

Local target: The local target option has the lowest deduplication potential as it uses larger blocks of data for the deduplication algorithm, but it also has the least CPU overhead.

LAN target: The LAN target option implies that the Veeam job definition points to a remote resource such as a mapped drive or NAS appliance. This deduplication algorithm for LAN uses a block size half the size of the local target option, with slightly higher CPU overhead to implement the deduplication algorithm.

WAN target: The WAN target optimization uses a deduplication algorithm with blocks half the size of the LAN target. Again, to increase deduplication, it will reduce the amount of transfer for the backup or replication job, albeit with increased CPU cycles to implement the most granular option within these options.

The WAN target option can add deduplication efficiency to the incremental jobs. Check out this recent Veeam forum discussion where one user identifies storage savings with the WAN target option.

Using the deduplication and compression features together can make the Veeam Backup & Replication files for backup jobs as small as possible. If backup jobs are organized from the originating template within a folder and like operating system(s); the optimization potential increases for deduplication and compression.

This functionality is explained in more detail on page 14 of the User Guide of Veeam Backup & Replication.

SureBackup and vPower

We started off the Spring of 2010 with our big announcement regarding SureBackup. This announcement was important for us because we knew that it was going to take some time to understand just what Recovery Verification was all about.

When you introduce something new to the market, you need to give enough time for it to soak in. When the summer came around, we started talking about vPower or Virtualization Powered-Data Protection. I gave the first live demo of SureBackup Recovery Verification, Instant VM Recovery and Universal Application Item Recovery (U-AIR) at TechFieldDay 3 in Seattle and it was very well received. So well received in fact that one of our competitors announced something very similar to our patent-pending Instant VM Recovery a few weeks later that to this day still has not been released.

Microsoft Partner of the Year Finalist

At Microsoft’s Worldwide Partner Conference held in June, Veeam was recognized as a finalist for Partner of the Year in the Core Infrastructure Solutions, Systems Management category. This was great recognition for Veeam, specifically around our Veeam nworks Management Pack for VMware.

Launch of Veeam ONE

Veeam is more than Backup & Replication, we also have a major focus on VMware infrastructure management. Veeam ONE was launched to create a single solution that customers could look to for monitoring, capacity planning, change management, reporting and chargeback. Also, since different customers need different monitoring solutions depending on their environment, Veeam ONE comes in 3 different flavors, including flavors for Microsoft System Center and HP Operations Manager.

Best of VMworld

For me the best part of VMworld 2010 in San Francisco was winning 4 “Best of VMworld” awards. Not only did the then current version of Veeam Backup & Replication (v4) get a finalist award in its category for the 2nd year in a row, Veeam ONE also was chosen as a finalist in its category. Of course when it comes to new technology, Veeam Backup & Replication v5 was the big winner with the Gold award and the top prize, Best of Show. This was the first time in the history of the awards that one vendor received 4 awards.

v5 Launch

With all of the excitement and buzz we created around SureBackup and vPower, it was great to finally release v5 to the market and watch it take off. For the launch “event,” we had more than a thousand people on a live webcast and on that day we hit an all time record for web traffic to veeam.com. So far we’ve had over 60,000 downloads and so many positive comments regarding the revolutionary functionality.

Informal poll at Gartner Data Center Conference

While not an official Gartner survey, this quick poll of 178 attendees at the 2010 Gartner Data Center Conference shows that Veeam Backup & Replication is used by the same percentage as backup solutions from the “bigger” vendors. Here’s a graphic showing the results:

The Awards!

There has been no shortage of awards for Veeam in 2010 and they seem to keep rolling in. The following were all in addition to the VMworld awards mentioned above.

• Best Free vSphere Tools – As I previously blogged about, Veeam took the top 3 spots in the VMwarevideos.com and vChat Best Free vSphere Tools survey.
  
• Virtualization Review – Veeam was also awarded 2 Reader’s Choice Awards from Virtualization Review in their 2011 Virtualization Buyer’s Guide. Veeam ONE took a merit award and Veeam Backup & Replication v5 was named a winner. These awards are based on reader input so we would like to thank everyone who voted for Veeam.
  
• SearchServerVirtualization.com Products of the Year 2010 – We were very happy to have won 2 Product of the Year awards from SearchServerVirtualization.com. Veeam Backup & Replication v5: 2010 Product of the Year in the Backup and Storage for Virtualization and v4 was also the Product of the year for 2009 in the same category – that’s 2 years in a row! Also, Veeam ONE: 2010 Product of the Year in the Virtualization Management category, that marks the first time in the history of the awards that one company has taken GOLD in 2 categories in the same year.

  

• Storage Magazine | SearchStorage.com Product of the Year – Veeam Backup & Replication v5: 2010 Product of the Year in the Backup and Disaster Recovery Software and Services category. When looking at who we beat out, this is a pretty significant award for us!
  

The Results!

155%

Of course all the awards don’t really mean much without good financials to back them up. As our press release announced, Veeam had a very, very good year financially with 155% growth over last year. Being a private company we don’t detail the exact revenue, but when you consider we’ve grown over 100% each of the past 4 years it’s not difficult to figure out. So when you consider Veeam as a company to do business with, understand that we’re still growing, we’re still 100% self-funded and we’re profitable!

Backing up a VM template will function a lot like backing up a powered-off VM (See my recent post on backing up powered-off VMs). Within Veeam Backup & Replication v5, there is an option to switch the view to include templates in the display, as shown in the figure below:

This type of object will occur much like any other backup job with the notable exception that a snapshot is not performed. Secondly, the template VM backup will be quicker to skip if there are no changes. Likewise in the restore wizard, it is generally the same as what would occur on a normal VM. The only caveat is the Instant VM Recovery feature does not work the same for a template as it would a normal VM. The primary reason for this is that a template cannot be powered on. This leaves individual file recovery and full VM recovery as options for the templates. There also is not support for changed block tracking (CBT), so each change is picked up as a full backup.

The default behavior for backup jobs is to exclude VM templates, and that is an easy option to change within the exclusions options of a job as shown in the figure below:

In most vSphere environments, there would be a handful of templates. Maybe one for every edition of the operating system, and possibly some pre-built operating systems with applications installed in them for frequently deployed configurations. In the case of Veeam Backup & Replication, template backups have been supported for a while, going back to version 3.

How have you addressed data protection for templates? Share your comments below.

However, there are still questions that may arise for each of the virtual machines behind the proxy appliance (located at the top end of the diagram). Here are five facts about these machines behind the proxy appliance:

1. The SureBackup virtual machines cannot pass through the Proxy Appliance to the production network. The VMs that reside in the isolated network are on a vSwitch that is only accessible to the proxy appliance. There is no route from this network out other networks.
With this safeguard, the isolated network virtual machines have the same IP addresses as the production virtual machines. This is the fundamental way that the virtual machines can start as expected while not interfering with the production instances within the isolated network.

2. The Virtual Lab proxy appliance is created from the wizard. When the Virtual Lab wizard creates the networking configuration for the proxy appliance, the virtual machine that performs the proxy appliance role is created with the input from the wizard. The proxy appliance is created from the contents of the C:\Program Files\Veeam\Backup and Replication\LiveCD folder in default installations. Each proxy appliance will have individual network configuration from the Virtual Lab wizard , which creates a floppy image to provide the proxy appliance its configuration. The virtual floppy image also is where logs are written. The floppy image file is stored as config.flp file in proxy appliance folder on selected datastore.

3. The vSwitch has no physical adapters assigned. If you look in the network inventory of the host, you may notice that there are no vmnic interfaces assigned to it. The figure below shows an example vSwitch for a Virtual Lab:

This is, by design, to keep the virtual machines in the Virtual Lab isolated. The SureBackup proxy appliance does have a connection on a production network vSwitch, but its configuration prohibits pass-through from the isolated to production networks. Inbound traffic is permitted through the access IP, if configured, within the Virtual Lab wizard.

4. The vSwitch is located only on one host. During the Virtual Lab wizard, the selected host is the only one that receives the vSwitch for the SureBackup Virtual Lab. In this way, the Virtual Lab is contained to this host and will not move around should a vSphere feature such as vMotion be configured.

5. Distributed Virtual Switches (DVS) are supported with Veeam Backup and Replication 5.0.1. Using a DVS on the ESX host is now supported for the Virtual Lab. Further hosts that are part of DRS clusters are also supported with version 5.0.1.

The Virtual Lab is an incredibly useful tool, yet the networking configuration should be fully understood before proceeding. If you haven’t checked out the User Guide, it’s the authoritative resource for these feature. Some lab time to go through the drill with vPower is also a good idea to be fully comfortable.

How do you use the Virtual Lab within Veeam Backup and Replication v5? Share your comments below.

In the Reader’s Choice Awards, Veeam Backup and Replication v5 was selected as the winner of the Virtual Business Continuity category and the Veeam ONE Solution for VMware received the Merit Award in the Virtual Management & Optimization category.

This week, the Americas Veeam team is gathered in Atlanta for the annual kickoff for 2011; and this is great news for us to celebrate together. While we are logistically hampered by strong winter weather in the Southeast, this is a bright spot and a big deal for the Veeam team.

The Reader’s Choice Awards are available online at Virtualization Review (registration required) and are also figured in this month’s print issue of Virtualization Review magazine. Be sure to check out Veeam in the Reader's Choice Awards. Thank you virtualization community!

While DPM does a good job of managing hosts during periods of low activity within a cluster, it doesn’t do much for virtual machines. There can be plenty of situations where virtual machines could be subject to being powered off during out-of-service hours. Shutting down Windows Servers automatically is easy enough with centralized scripts using PowerShell, shutdown.exe or even a local scheduled task. The revival task to power on the virtual machine is also an easy one-liner that can be easily automated with the Start-VM PowerCLI command, as shown below:

If a number of virtual machines are permitted to be powered off, the use case for DPM becomes more attractive. While idle virtual machines will release memory back to the host via VMware’s memory management techniques, the ultimate resource giveback is a powered off state. To be fair, when the virtual machine is powered back on, DRS and some of the memory management techniques may take time to re-arrange all of the newly added resources to the cluster.

Regardless of the reason a virtual machine is powered off, Veeam Backup and Replication can still protect the virtual machine in this state. In fact, the snapshot burden that the virtual machine, host and storage systems undergo is lessened by the fact that the redo-log files have effectively nothing to write to the source VMDK. While the virtual machine is powered off, features such as VMware’s Changed Block Tracking of each VMDK are still utilized during the backup job for optimal performance. Further, all of Veeam's restore and vPower features are available for the powered off virtual machine.

There are two important notes for backup up powered-off virtual machines, the first is that application-aware processing does not occur on a powered-off VMs and the second is that this is not a good use case for a domain controller. See this earlier blog post about what may happen for a domain controller.

As a side note, if multiple backups are run on the virtual machine while it is powered off; the subsequent backups (when using incremental mode) will be a small container only of headers because nothing has changed. In regards to excluding powered off virtual machines from a backup job, this was raised in a forum discussion and the current product only allows exclusion of templates from the backup job.

Most agent-based backup tools can't back up powered-off virtual machines. This may be an obstacle for taking DPM to the next level and setting up automated shutdown and power up operations for virtual machines, yet ensuring they are backed up. Do you perform automatic shutdown operations of virtual machines during out of service hours? Would you if you could back them up? If so, explain your environment below.