Using iland object storage for Veeam capacity tier
If you have been in the IT game long enough, you have likely heard of object storage. Amazon S3 seems to be the gold standard as far as storage object storage protocols go as many 3rd party object storage vendors claim to be S3 compliant. All the large hyperscalers have an offering, Azure Blob Storage and Google Cloud Storage are the offerings from the big 3. So what is Object Storage? Well like you, I can grab the Wikipedia article which states that it is a system that allows storing massive quantities of unstructured data. All data is placed as an object (hence the name) and is typically accompanied by some metadata which is effectively a pointer and a description of what the underlying data in the object is.
So, why would you not just consume object storage from one of the aforementioned hyperscalers? Cost is a big part of that question. If you check out this guide from Amazon regarding object storage action types, there are quite a few. Each action performed typical incurs a small cost to operate. When writing and retrieving lots of data, those costs can quickly stack up.
Now with an object storage provider like iland, all you pay for is the cost per TB of storage required.
In this article, I am going to cover how you can leverage object storage to create a second copy of your primary backup data by utilising Veeam capacity tier functionality.
iland object storage
Most services offered by iland can be managed and monitored via the iland console, object storage included. One of the first things we need to do is create secure access keys to the object storage and create a new storage bucket. A bucket is a location where the data written to object storage will reside.
Log into iland console and navigate to the object storage section. Click the ellipsis in the buckets section and choose to add new.
Define a suitable name for the bucket.
To define access credentials, click on the settings button and then generate new credentials.
Make a note of the Acess Key and Secret, you will need these later to access the storage from within Veeam.
Also, make a note of the Object storage URL, this is the target you will connect to from within Veeam.
Lastly, set the bucket to be private so that only you can access it. Public has its uses, but not as a Veeam target!
Adding object storage to Veeam
You now have all you need to be able to add the object storage to Veeam. Object storage in this instance is leveraged as a capacity tier as Veeam dubs it which forms part of a Veeam scale out backup repository. A scale out backup repository, or SOBR is a culmination of many backup targets, presented as a single destination for backups to write to. The benefit here is that backups can write simultaneously to many disk extents, increasing backup throughput or some logic can be applied to write full backups to dedupe storage for example but incrementals to quicker JBOD or RAID disks. The object storage drops into the scale out backup repository to offer one of two things. A mirror image of the backups that reside on the local backup targets or a destination to send older backups to once they are no longer required on the local performance tier. This makes sense because really, most data restores are from the most recent or very recent backup sets. This is when you want to have the backup available on nearline storage for quick recovery. After a certain time frame, where recovery from older backup sets is less likely to happen, then it makes sense to move those backups some where else to save space on the more expensive, faster storage.
In Veeam, navigate to backup infrastructure and then add a backup repsoitory.
We want to add object storage.
As I mentioned at the start of this article, most object storage uses S3 compatible protocols to interact with the storage. Choose S3 compatible.
Give the object storage repository a name.
The URL I said you would need to grab at the start of this article, pop it into the service point box. Type in a name in the region box and if you have not already done so, define the credentials to be used which are the access key and secret.
Once a connection has been established, click on browse to find the bucket you created earlier.
And for the folder, click browse and create a new folder. I also limited the storage to the amount of storage available. if you were doing this in a hyperscaler where you could potentially write as much data as you want, this would be a good way to limit that and have a more predictable spend.
The iland success center article suggests creating the following registry key to lower the maximum number of tasks but to increase the storage optimisation on the backup job to local file 16TB+. There is a good explanation of what this registry key does in this Veeam forum thread. TBC if still relevant with VBR 11
Configure the SOBR
I have made an assumption that you have already created a Scale Out backup Repository, this needs to be in place before you can make use of the capacity tier. Open the properties for the SOBR.
For this example, I am configuring the Capacity tier to copy backups immediately to object storage, effectively a mirror image of what resides on the performance tier. Be mindful that this option effectively doubles the storage requirements for your primary backups, but it does have the benefit of creating an offsite copy of the data.
With the copy mode selected, any existing backups on the performance tier will be queued for copy, so bear this in mind if you enable the option as it may take a while to complete.
Once the backups are copied, you can check them out in your bucket in the iland console. The folder name you created earlier is where the backups will reside.
Now you have another copy of the data, how can you recover from it? Well, it is very simple and use the same familiar recovery process used when recovering from any media type.
From the Veeam Backup and Replication server that is integrated with the object storage, the backup location appears as like any other with the same recovery options
Now let’s imagine the primary Veeam server has died for some reason, how do you recover from those backups? You can import the object storage back into Veeam and recover from the imported backups as covered in this topic on the Veeam website.
Object storage is another great resource to help build out a Veeam infrastructure to meet the 3-2-1 backup rule, 3 copies of data, 2 different media types and 1 offsite. Integration with iland is easy and there are no hidden costs to be concerned with.