Aller au contenu principal

Multiple backup jobs Hyper-V Cluster

Thread needs solution

I am running Windows Server 2012 r2, Hyper-V Failover Cluster with 3 Nodes. Currently I have backup jobs setup for multiple VMs hosted inside the cluster. All the VMs are stored in Shared Cluster Volume. However, each backup jobs seems to run multiple times (once per Node with acronis agent installed). Is there any method which I am able to set it to run only on the Node the VM is residing or only on 1 of the Nodes? 

Thanks.

0 Users found this helpful
frestogaslorastaswastavewroviwroclolacorashibushurutraciwrubrishabenichikucrijorejenufrilomuwrigaslowrikejawrachosleratiswurelaseriprouobrunoviswosuthitribrepakotritopislivadrauibretisetewrapenuwrapi
Contributions: 22
Commentaires: 3800

Hi Jordon,

In case you're selecting a clustered VMs for backup (under Hyper-V cluster in GUI) the backup plan will be applied to all nodes of the cluster and will finish silently on the nodes which do not host the backed up VM currently, while proceeding normally on the active node. This is a normal behavior for centralized backup tasks which is designed to track and handle VM migration events (so VM is always protected regardless of its current location).

If you want to apply backup plan to a single Hyper-V node only, then it has to be personal backup plan which can be created if you connect to specific Hyper-V node directly from Acronis Backup management console from "Machines with Agents" section (right-click -> connect directly). Note that in this case if the backed up VM migrates from this node then the backup plan will fail and VM will not be backed up.

Thank you.

Hi Vasily, 

Thanks for the reply.

"If you want to apply backup plan to a single Hyper-V node only, then it has to be personal backup plan which can be created if you connect to specific Hyper-V node directly from Acronis Backup management console from "Machines with Agents" section (right-click -> connect directly). Note that in this case if the backed up VM migrates from this node then the backup plan will fail and VM will not be backed up."

Not ideal in my enviroment, because the VMs will move to other Nodes when maintenance is done on the physical server.

 

Is there any other settings in which the backup job will run only on the active node the VM is residing?
     Example:
                   Node1 has VM_01 running on it
                   Node2 has VM_02 running on it
                   Backup will only run on Node1 for VM_01 and Node2 for VM_02

      If VM_01 move from Node1 to Node2,
                   Backup will not run on Node01
                   Backup will now run at Node2 for VM_01 and VM_02.

Reason for this is due to huge amount of disk space and system resources is required if the backup is running on every Nodes in the cluster.

frestogaslorastaswastavewroviwroclolacorashibushurutraciwrubrishabenichikucrijorejenufrilomuwrigaslowrikejawrachosleratiswurelaseriprouobrunoviswosuthitribrepakotritopislivadrauibretisetewrapenuwrapi
Contributions: 22
Commentaires: 3800

Hi Jordon,

Probably I haven't described it properly and this caused confusion. My main point is that while the backup tasks are deployed to both nodes and executed on each of them at the same time, it is normal and does not cause any additional impact on the nodes. At the time of backup the actual backup activity is performed _only_ on the active node where the VM currently resides. On the other "inactive" node the backup task runs, checks that the VM is not there and silently succeeds without doing anything, thus not consuming disk space or system resources on "inactive" node. If you see some different behavior then it's a problem which should be investigated separately (via logs/screen shots).

Thank you.