Changing VM backups

[Tom84]

Hi

When I first set-up Veeam I used the entire server as the backup source, instead of the individual guests. I have two hosts which replicate to each other every night, this set-up obviously causes problems as if I don't exclude the replica it will then replicate the replica and append another replica to the source (vicious circle!).

So my questions are:

1. If I change the "virtual machines to backup" from the hosts to the individual guests will this disrupt anything? I don't want to lose any previous replicas or backups, and I don't have the space if the job decides to disregard the old replica or backup file and create a new set.

2. Side question, on the guest processing settings, if the guest is listed in both the "host" and added as an individual which setting will it use?

Just to clarify #1, I want to change an existing job from listing the hosts:



To the guests:



And for #2, TS1 is a guest within 192.168.1.12, so which setting would it use?



Thanks

[TrevorBell]

Hi ,

Do you not have Vcenter installed or do you just backup host to host ?

[Tom84]

Yes, we have Vcenter installed, but we're just backing up host to host. They are two hosts with DAS, there is no SAN.

Is there another method I should be using?

[TrevorBell]

For sure...

Use Vcentre its much better, add Vcentre then you will see your two hosts under the Vcentre Icon and then setup your backup jobs again. ( or add another job so yo do not lose the job data you already have ) and then after the retension period of the "old" backup just use the vcentre backups

does that make sense ?

[Tom84]

What advantages are there to adding the guests/hosts via VC? I already have VC added to the Veeam console (alongside the hosts).

I understand your method, but unfortunately as stated in the 1st posted I don't have enough space to retain two sets of backups / replicas.

Which brings us back to question #1, what happens if I just edit the existing jobs and change the source? (as seen in the first two pictures above).

[Vitaliy S.]

The advantage is that when you add VMs via vCenter Server then your job becomes "vMotion aware", in other words when any VM migrates from Host A to Host B then it will be processed from scratch as a new one. Please note that you can add hosts via vCenter Server as well as doing it in a standalone way.

As regards your first question, then my answer depends on your setup configuration. As you might know all VMs are processed with the reference to their unique ID assigned by the host or vCenter Server. That said, if you've added hosts via vCenter Server then all your VMs will have the same ID and you can safely change hosts containers to individual VMs adding them via vCenter Server.

On the other hand, if you've added hosts as standalone to your backup console, then all VMs will have different IDs and adding them through vCenter Server will cause a new full run.

I haven't tried that myself, but I believe you should stick to your initial setup - add VMs through standalone hosts, so you could keep their ID and continue running incremental passes.

As for the second question, then VM's requirement will prevail in the job configuration.

Hope it helps!

[Tom84]

Thanks.

That makes perfect sense and if I understand you correctly, because I've been selecting the host as the current source, I should be able to "open up" the host and select individual VMs without the ID changing but if I chose to backup individual guests via vCenter it would. I may end up keeping my current solution, the "replica_replica_replica" circle isn't that much of an issue as I'm not adding new guests that often.

We aren't using vMotion- but good to know.

Regards #2, that's great and the behaviour I was hoping for.

[Vitaliy S.]

Yes, you got my point correctly, drilling down the standalone host should make a trick.