Some time ago, one of my friends had a house fire. They literally had only enough time to grab the kids and run for their lives. The fire burned so hot, and so fast, they could only run. And their car, parked at the curb, had its house facing windows melted. They lost everything. My wife and I travel the world, and have been doing that for a long time. The pictures we have are breathtaking. For example, a howler monkey crawled on me, and fell asleep once. Another time an old bull Elephant took my measure with his trunk only a few inches from my face while he – so tradition goes – memorized my scent. My wife and I, along with a few others were attacked once by Army Ants. These pictures cannot be lost. After Pete had his adventure I figured out how to do things different. Pictures and other important data have a file backup (via Time Machine), and image backup (using SuperDuper ) in the house, and everything is also backed up in the cloud (BackBlaze).
So this take care of my personal stuff – I can handle deleting a file, losing a hard drive, or losing the house. But what about in the work side of things?
I know several Veeam customers that do backups to the Veeam server and that takes care of quick restores, and they back up to an SMB device that that takes care of a little more serious issue like losing the Veeam server. But that is all they do.
This article is about adding in that very important offsite copy. Veeam Cloud Connect is easy, and in fact inexpensive (one of my ‘bosses’ said he knows one Service Provider that charges 49 US$ per TB per month so that confirms not expensive I think). I will show you how to do it below.
I am using OffsiteDataSync – thanks to Hannah Coney (many thanks actually!) – which has turned out very well – I quite like their UI, and their product.
So let’s get started!
What do we need to have handy?
- We need the FQDN or IP address for the Cloud Connect service provider.
- We need credentials to use the Cloud Connect service.
- We need to know which VMs we will back up to the cloud.
- We need to have admin access in Backup & Replication.
Configure the Service Provider
- We need to change to the Backup Infrastructure View, and select Service Providers. You can see what this looks like below.
- Now we have a Service Provide wizard to help. First question is the DNS or IP info for the service provider – which the welcome email or their UI will provide.
- I love the option – that we are not using – about letting the service provider to manage my Veeam instance – not for me but I bet there are a bunch of people that do like it. Once you select Next you will connect.
- We can see we connected, and we can check the cert info and it all looks good.
- You need to use the Add button to add your Service Provided supplied credentials. Make sure to get them right, as you won’t get past here if you don’t.
- Once you get the creds right, and get past this screen you see the resources that the service provider has allocated for you.
- In my case, I have 512 GB to work with and only one repository. I don’t have a huge lab, so 500 GB is great – I can protect it all. Also notice that WAN is enabled? This is very cool. I will chat about that later. You need not make any decisions on this screen.
- We see the log of the relationship between your Backup & Replication server and the cloud. But you can hit Next to continue – when you can.
- Now you are done. You have one additional repository now – of type Cloud.
- What is important to understand, is that this cloud based repository is the ‘same’ as the Windows ones we see above. This means each of them, any of them, could be used for Backups, or Backup Copy jobs (or restores too).
Getting your backup offsite
I have backups configured now that I like. What I want to do now is to get my Active Directory VMs offsite. So let’s do that!
- I love the ability to only impact a production virtual machine once. And then, for other replication and backup copy jobs we can use as a source the backup and not the original VM.
- So let’s create a Backup Copy job. This is what we will use to move our backup into the Cloud – without impacting the original VM.
- Now we get a wizard to help.
- The Copy every: we see above is sometimes a little confusing for people. Basically this means the Backup Copy will watch the backup repository, and when a new backup finishes, it will be copied once.
- We now select the virtual machines for our Backup Copy job. Generally, use From Backups, but the thing to remember is if the VM you are selecting is not yet protected, you should select if From Infrastructure.
- Since I want to protect my AD infrastructure I select the AD job above, and the VMs from within it.
- Logan, Logan2 are the names of my DCs which explains why it is seen in the list above!
- Now we see a screen we have seen before. In fact, it is all normal – right? Even though this repository is in the clouds! As you see above I like to keep 2 (default is 7) restore points, and in Advanced I do set my notifications too. I like to manage my Veeam via emails! So very minimal involvement on my part unless more is required.
- I did not have a WAN accelerator at first. But it turns out it is very easy to add and so I did. It does make quite a difference for the upload to the cloud in that it makes it much faster, and uses less bandwidth too. I will talk a little more about this below.
- Defaults is fine in our case of backing up the DC’s to the cloud as we are using backup jobs as our source and not impacting end users.
- I leave the check-box enabled as who wants to wait?
- We can see we have a new job – BC_AD_Cloud – and it has a target of the cloud. Nice!
- See below what it looks like. Note how it looks like any other Backup Copy Job. Exactly as it should!
- And we get a nice email about the job too.
- While we have a WAN accelerator in use, it has not been seeded. If the destination WAN accelerator was on our remote site we could have seeded it and our first backup copy job would have gone much faster. In our case here we are seeding the upstream WAN accelerator as we do our first Backup Copy Job. But it does finish eventually! The next time we do a Backup Copy Job with VM(s) that are also Windows the upload will be much faster due to the WAN accelerator being seeded.
- You can use this link to find your own service provider, and make sure to pick someone that is in your general area.
BTW, what you see at the cloud provider – in this case OffSiteDataSync, is pretty clean.
WAN acceleration background
If you have two sites and are doing a Backup Copy Job between them that is a good idea – gets your backups off site. A better one is to have a WAN accelerator in both sites. You will own both of them so can do seeding which will allow great performance and bandwidth savings. But in our case here I do not own or have control of the destination WAN accelerator so I have to seed it ‘naturally’ which is the least efficient way to do that but it could still help on the next upload. You can learn more about WAN acceleration in this set of articles.
Deleting backups out of the cloud
Just in case you need to delete a backup that is up in the cloud you can do it from within your Veeam console.
Now you can right + click and select remove from disk.
We have configured a service provider connection, and we have created a repository that is actually in the clouds, and have configured a backup copy job to use it. Now we have locally in the lab a Windows repository in the Veeam VM, and one on SMB repository on an external NAS, so that means I can handle a lost SMB drive or the whole NAS, or a lost VMDK – from the Veeam VM – or the whole VM, and still be able to restore from our backup in the cloud so we can even handle losing the office!
An important note to remember is that if your backups are up in the cloud, then any CryptoLocker type issues will not be able to manipulate them.
- 9/17/16 – Luca published an article recently that shows how you can seed your Cloud connection. That means if you have slow links between your B&R or WAN accelerator machines and the Cloud this can help you get that first very big replication done MUCH easier.
- 9/9/16 – updated screenshots, and added Hannah’s name! Very big thanks to OffsiteDataSync and Hannah Coney for the offsite repository.
As always, comments or questions are most appreciated.
=== END ===