Host-based backup of VMware vSphere VMs.
Post Reply
bishoptf
Enthusiast
Posts: 25
Liked: never
Joined: Oct 06, 2011 2:27 am
Full Name: Tom Bishop
Contact:

Backup Copy Jobs to rotated media

Post by bishoptf »

Still trying to get my head wrapped around how this should be setup and how it should work. I have searched and read most of the postings but I'm still not fully understanding how I should set it up in my environment.

Here are my details and what I am trying to accomplish. I have a Linux NAS that is also my veeam repo, my normal backup jobs run nightly and hit the linux nas repo. Within that same linux nas repo I have to hot swap bays that I will be rotating disks offsite, the main veeam backup jobs are written to the raid array nightly.

What I want to do is to replicate the backup information from the raid array to my hot swap bays, then weekly I will be rotating the drives. I see how to create a new repo and create a backup copy job and enable rotating drives option. My concern is that when I get to the point about mapping the directory on the repo, it uses the uuid for the mount point and I assume I would have to make a repo for each drive that I rotate in and out. One of my concerns is what happens to the backup copy job that has been created for a particular mount point and is no longer available when the drive is offsite.

Not sure if this makes any sense or not but my concern is how to backup copy jobs handle when the repo using the disk are not found. I assume I will get an error message every night until the drive comes back into the rotation.

If anyone has any insight into how this should work or how I should set it up please let me know.

Thanks
PTide
Product Manager
Posts: 6408
Liked: 724 times
Joined: May 19, 2015 1:46 pm
Contact:

Re: Backup Copy Jobs to rotated media

Post by PTide »

Hi,
Still trying to get my head wrapped around how this should be setup and how it should work
If rotated drives option is enabled on repo then any backup or backup copy job checks if the chain is consistent. If anything is missing (drives swapped) the chain is started anew.
how to backup copy jobs handle when the repo using the disk are not found
Backup copy job does not look for drives, instead it looks for files in a directory. Even if your drive is not mounted properly the mount point is still present, which means that you backup copy job will still be able to write into that directory, no matter what's behind it, resulting in writing to the wrong device in your case. Assuming you have 5 hot-wap drives with uuids 1,2,3,4,5. Can you set up fstab on your NAS in such way that it mounts any of those uuids to the same point, say, /root/mountpoint/rotated ?

I'm not sure that I got you right, so please let me know if I misunderstood something.

Thank you.
t_ahrens
Influencer
Posts: 13
Liked: 1 time
Joined: Feb 17, 2016 12:50 pm
Full Name: Thorger Ahrens
Location: Hamburg, Germany
Contact:

Re: Backup Copy Jobs to rotated media

Post by t_ahrens »

Hi,

I'm handling with almost the same issue, but I don't want a new full backup to be created every day, because it would take to long.

Isn't it a possiblity, to set up two or more Backup Copy Jobs with different names like Copy1, Copy2 and so on map them to the same CIFS/USB drive (not the same disk and not at the same time, but one after the other)?
So each Job will work, while the right harddrive with the assigned folder is in the bay (or externally attached to the server, should be the same scenario).
And if another diks is insertet / attached, the job will throw an error like "path not found". When the right disk is inserted /attached again, the Copy Job starts an produces the next increment.

Or is ther a mistake in my thoughts and this would cause problems (besides the error messages)?

Greetings, Thorger
foggy
Veeam Software
Posts: 21069
Liked: 2115 times
Joined: Jul 11, 2011 10:22 am
Full Name: Alexander Fogelson
Contact:

Re: Backup Copy Jobs to rotated media

Post by foggy »

Thorger, two jobs can work, provided you do not enable rotated drives option for the repository. However, a more preferred way of doing this is enabling it and using a single job: if it is a Windows-type repository, the job should maintain a separate incremental chain on each of the drives.
Post Reply

Who is online

Users browsing this forum: KonstantinS and 88 guests