Orchestrator options for in-guest iSCSI

[mjm599]

Hi All,

We have some Windows VMs that have in-guest iSCSI configured from a NetApp Array. Current DR Solution consists of DR software that replicates the VM to the DR Site. We then also have NetApp SnapMirror to replicate the NetApp Storage where in-Guest iSCSI volumes reside to DR. So currently, to failover the VM with the existing DR Solution, we failover the VM to DR, then break the snapmirror and bring volumes online in DR so they can then be presented via in-guest iSCSI to the failed over VM.

The new solution is Veeam B&R aswell as Recovery Orchestrator so we are looking to see how we can address the above for those VMs with in-guest iSCSI. Looking at this: https://helpcenter.veeam.com/docs/vro/u ... tml?ver=70 it appears this wont cater for in-guest iSCSI to VMs. So i know there is the custom script option in VRO, would it be a case of adding a custom script in vRO so that can then execute relevant commands on the netapp as part of failover?

Thanks all

[Alec King]

Hello,
You are correct, the VRO integration for NetApp ONTAP is designed to failover the storage where VM disks are stored, then to re-register the VMs in the new location. If the storage is used solely for in-guest iSCSI then it would have to be handled differently.
You could run a pre-plan script step in VRO that will break the snapmirror relationship and failover to the replicated volume.

Are you planning to use a Veeam replication job to copy the VMs to the DR site? If so, then you can use a VRO replication plan based on the same job. The main part of the plan will handle replica failover for the required VMs, and the pre-plan steps can handle the storage failover using scripts.

A tip: VRO can pass in many runtime parameters to a script, so for example if you pass in the parameter %plan_state% then your script can differentiate between failover versus failback, and perform the relevant storage commands without human intervention.
Another tip: VRO can execute planned failover by using some built-in pre-plan steps - you can trigger the VM replication job, then power VMs down, trigger the job again, then execute storage and VM failover and power on all VMs. That will perform a full planned failover with zero data loss.

Hope that helps!

[mjm599]

Hi Alec, thanks for the reply and details. The VMs will use Veeam CDP to be replicated from Site A to Site B. So we have some VMs that do not have in-guest iSCSI and then some VMs that do have in-guest iSCSI.

[Alec King]

OK, then my suggestion above would work -
- Create a CDP Replica plan in VRO (you can use the group we auto-detect based on the VBR policy)
- Import your PowerShell script to VRO as a custom step, and trigger it in pre- (or post- ) plan steps to manage the storage failover.
- You can also run scripts inside each guest, if any remapping of iSCSI targets is needed.

Any further questions, just let us know. Thanks!