It’s finally there: Veeam Backup & Replication 9.5 comes with many new features and improvements around the corner, sweetening the last weeks of every backup admin before Christmas  😉

 

I’ve already done some testing in our very basic environment and I really love the improvements so far!

One of the most interesting features for me is the integration of ReFS 3.1 from Windows Server 2016. With ReFS 3.1 comes a a set of improvements on file-system level, these are the most important for me:

  • ReFS uses data integrity streams to ensure that the data itself won’t get corrupted – a 1 stays a 1 and 0 stays 0 and won’t become a 1
  • Veeam introduced the fast clone technology which improves / shortens the creation and transformation time of a synthetic backup by the factor 20! I tested it only with three reversed incremental jobs, but the difference from Server 2012 with Veeam Backup & Replication 9.0 Update 2 to Server 2016 with Veeam 9.5 is enormous!In order to make Veeam 9.5 use the features of ReFS 3.1 the backup target has to be formatted under Server 2016. Server 2012 has also been able to format partitons with ReFS, but that version of the filesystem didn’t have those improvements. The backup repository has to be created with 9.5, upgrades to 9.5 won’t upgrade the repository. Also you have to to an active full backup to the new repository, only copying your old backups won’t do the trick.

How do you know if Veeam 9.5 makes use of ReFS 3.1?

After creating the new repository have a look at the details of the repository under „Repository“.

If „Align backup file data blocks“ is greyed out, then 9.5 makes use of the enhancements of ReFS 3.1

If "Align backup file data blocks" is greyed out, then its a ReFS 3.1 repository
If „Align backup file data blocks“ is greyed out, then its a ReFS 3.1 repository

 

Did you ever wonder why it takes so long when you click thru the lists of your VMs in Veeam?
With 9.5 you don’t have to wait any longer! The GUI doesn’t take so long to present you a list of your VMs and Veeam promises also that it accelerates the start up of jobs, as the time for building VM list operation will be reduced. What I can say so far is: yes they are right! The time decreases in my tests by the factor 3-4 for the VM list operation and listing your VMs is also faster.

Speaking about performance, they have now enabled parallel processing for full VM restore. Great! Especially the restore of VMs in the case of disaster is time critical and Instant Recovery  (which comes along with more performance, who wonders?) is not always an option.
Parallel processing works for all disk based repositories, except Data Domain.
And not disk based repositories like tape? You are now able to recover a full VM directly from tape without restoring the complete backup to another repository first. An improvement you will love when it comes to a crash for the first time and you need your tapes, but can’t do a restore due to the lack of space…

But there is more, much more!

  • There have been many improvements for cloud users,
  • or users of Veeam Agent for Windows who can now do a direct restore to Hyper-V,
  • the option to skip (instead of aborting the whole job) VMs if the free disk space of the production datastore is below x-% to prevent running out space due to open snapshots during the backup
  • and many many more. I advise you to read the official Veeam Backup & Replication 9.5 Whats New PDF

 

In another blog post I’ll present a few numbers on how the backup creation time has improved since I upgraded to Server 2016 and Veeam Beackup & Replication 9.5

Schreibe einen Kommentar

Deine E-Mail-Adresse wird nicht veröffentlicht. Erforderliche Felder sind mit * markiert.