Offsite backup options to replace ESXi and Veeam
-
Hey guys,
WIth the recent upheaval in VMware town i am looking at migrating a number of smaller customers from ESXi/Vsphere with Veeam for backups.
We typically backup to Local network attached media - either presented as SMB or NFS shares and then use the Veeam features to replicate the deltas in these backups off-site through some magic they do with CBT, synthetic backup and the like.
We find this works well and has saved us a couple of times with different customers and ransomware.
What options are there when using XO to move data to offsite storage ? Is there somewhere i can read up on these ?
As we are in Australia and Internet speeds are still somewhat limited (and typically asymmetric) pushing large backups offsite each evening is not really an option nor is waiting days for a restore to happen - so it would be good for generic storage options to be a target so we could host these for our clients (rather than an S3 style option)
Craig
-
@curto Hi Craig,
Unfortunately no option ro move the data directly inside XOA, but in your backup job you can define multiple "remotes" that are backup locations, so you could push your backup locally and remotely at the same time.
I think you should really try XOA it comes with 15 days trial that allow you all the features and full support if you need more just contact sales@vates.fr to extend the trial. -
Just to understand exactly, what you seek is doing backup of backup, in a async fashion, right?
-
@olivierlambert - Yes essentially we want to do a local backup to NAS storage on the customers site and then replicate this backup on a daily basis to a remote location - in the remote location we would like to end up with a Synthetic full backup so in the event of a full system crash at the clients premises we could load their backups onto a loan server and ship it back to site almost ready to go
Craig
-
@Darkbeldin - Thanks for this information. It is difficult to try it with most of these clients as they are single server sites (or at most two physical servers) - so we first have to spec out replacement hardware for them that is going to be adequate for whichever direction we take and then start trying to duplicate what they currently have.
I think i will get a setup in our lab and do some playing around and see where we can get to
Craig
-
We have "async" backup (backup of backup to another location) in the backlog. In the meantime, you can also rsync the content of the backup folder to another location to achieve a similar result.
-
Adding @florent in the loop
-
@curto hi,
also one of the specifity of our backup approach is that XO's remote ( = backup repository in veeam-speak ) are completly independant of the xo installation, no local block database. That means that if your XO is broken, you can spin up a new XO, add the remote connection information to the NFS , add the encryption key and restore backup.So if you rsync/rclone the remote to another location, you really got all the data needed to restart immediatly after an incident.
The goal is to be able to handle the cloning part directly in our backup scenario in a few month, please watch the annoucments as we will includes the community in the development, either in feature pruning or test in various environment.
Full remote encryption should be released in a few weeks after a test phase by the community, it's already working quite well in our lab.
Other targets that should be available in a few months : better handling of immutable remotes, AWS glacier and compatible, async cloning.
-
Good one Guys - thanks for all the info
We will spin up and instance and do some playing and see how close we can get using RSYNC - will follow along the blog for more info on when the remote feature set becomes richer as you develop
Craig
