Using iland object storage for Veeam NAS Archive Tier
Background
Following on from my previous blog post, using iland object storage for Veeam capacity tier, where I covered what object storage is and gave a primer into what iland object storage I take a look at how object storage can be leveraged as an archive target for Veeam NAS Backup. This blog is written when Veeam V11a is the current version, functionality may change in the future.
Veeam NAS backup works a little differently from how virtual machine backup works. One of the job types that does not exist in the traditional sense is a Backup Copy job. With a virtual machine backup copy job, you can do just that, create another copy of a backup in another location, including offsite at a Veeam service provider like iland. With Veeam NAS backup we have the concept of a secondary repository location and an archive repository location. If you check out this article from Veeam, we can see that a Veeam Cloud Connect repository is not supported at all, and object storage is supported as an archive repository. This makes getting NAS backup data offsite using one of those aforementioned services rather difficult.
How does the NAS archive tier work? Understanding this concept is key, Veeam has an article here covering various scenarios but in a nutshell, you need to understand that NO DATA will move to the archive tier unless the source data has either been modified or deleted. If you are thinking about using the archive tier as a secondary copy for the NAS backups, it will not work.
A picture paints a thousand words. Below we have a scenario where the primary NAS backup repository has a retention policy of one day. The Archive tier has a retention policy of one month. Two files are being backed up, File 1 is modified on day one, File 2 is not.
On day two, the archive policy kicks in for File 1 as it has been modified so the backup on day two for File 1 is version two of that file. As File 2 has not changed and is still version one, day 1 becomes the backup on day 2.
Hopefully, that makes sense!
Implementation
With the explanation of what Veeam NAS backup is and how it works out of the way, let’s jump into how to set it up.
Creating a NAS backup job
Adding an SMB share.
Naturally, if you want to backup a share on a NAS, you need to create a backup job to do so. In Veeam Backup and Replication, select to create a new backup and choose the File Share option.
If the file share to be backed up does not exist, you will be prompted to add one.
Click to add a new SMB share. It could be any of the options below.
Type the UNC path for the SMB share, provide credentials if required.
Choose a proxy to process the SMB share.
Wait for the addition to complete.
And done.
Create the backup job
Give the backup job a name.
Target the SMB share that was just added as the source of files to backup.
Define the number of versions of files to keep in the primary backup repository. We will come back to the archive settings later on in this guide.
The secondary target can be skipped in this scenario.
Define when the job will run.
And you are done.
Creating the archive tier
Iland object storage configuration
For this to work, we need an object storage bucket to where the archived backups will eventually be sent to. From the iland console, it is very straightforward to create a new bucket.
Log in, browse to object storage and click the create bucket option.
Give the bucket a name.
Then browse to settings to grab the URL and access credentials required to add the object storage to Veeam.
Adding object storage to Veeam
To be able to utilise the object storage as an archive tier, we need to add it to Veeam Backup and Replication. This is implemented slightly differently from an archive tier on a scale-out backup repository in that the object storage location is a standalone repository rather than forming part of the scale-out backup repository,
From the backup infrastructure section, add a new repository and choose object storage.
Choose S3 compatible.
Give the repository a name.
Enter the URL and access credentials that were captured earlier, to access the object storage.
Choose the new bucket that was created.
Create a new folder in the bucket. The folder can not be created in the iland console first, it has to be done in Veeam.
And you will see a new object storage repository in the backup repositories section.
Enable file archive on the file backup job
Circling back to the file backup job, we can now edit it to add the object storage as an archive tier.
Edit the job, and click through to the storage section. Here we can tick the box to keep previous file versions and define the archive repository as the object storage we just added to Veeam.
The settings above now match the scenario I portrayed at the beginning of this article.
Is it working?
There are two places you can check this. In the job statistics, it will state if any files have been archived and you can also check the properties of the backup set which should look a little something like this if any files have been archived.
Conclusion
Once you grasp the concept of how the archive tier works, it is pretty straightforward. Any unmodified source files will forever live in the primary backup repository. To be able to recover from previous versions of a file though, you can rely on the archive tier to retrieve that data. It is a nice concept because 99% of the time, the file to be recovered is most likely going to be the most recent copy of it, which will sit on a quick nearline backup repository for rapid recovery. But with the archive tier you have the additional safety net of being able to recover older versions of a file which could be particularly useful if the source files are subject to a ransomware attack. The primary files will have changed so the good copy of the backup in the primary repository will be marked as a candidate to move to the archive tier.