How Orchestrator Processes Resources
To restore VMs included into a plan to a recovery location, Orchestrator uses the following algorithm:
- The solution looks through all hosts added to the location as compute resources to detect the first available host. This is the host where the first processed VM will be registered.
- Orchestrator applies the mapping specified in the network mapping table for the location to set the required network configuration of the restored VM.
Orchestrator checks whether the network configuration of the detected host matches the required network configuration. If these configurations do not match, Orchestrator goes back to step 1.
- In the list of datastores connected to the host as storage resources, Orchestrator searches for the first datastore that both is available and has enough capacity. This is the datastore where files of the VM will be stored.
To calculate the datastore capacity and make sure that it has enough space to accommodate the restored VM, Orchestrator uses the threshold that you specify when creating or configuring the location.
When calculating the amount of free space available on a datastore, Orchestrator assumes that the swap file of the processed VM will be restored to this datastore as well. Orchestrator also takes into account that all disk files of the VM will be restored to the same datastore that stores the .VMX file, regardless of whether the source VM stores its disk files on multiple datastores. This may influence the resulting capacity estimation and the way Orchestrator searches for datastores to restore VMs.
The failure of steps 1–3 for a VM from a critical VM group halts the plan in the following cases:
To learn how to work with HALTED restore plans, see Managing Halted Plans.
- When Orchestrator starts processing the next VM, the solution looks through all hosts added to the location as Compute resources to detect the next available host:
- If Orchestrator detects such a host, this is the host where the VM will be registered.
- If there are no more available hosts, Orchestrator uses the host detected at step 1 to register the VM.
Then Orchestrator goes through steps 2–3 to detect the target network and datastore for the VM.
- Orchestrator repeats step 4 for all other VMs included in the plan until all the VMs are restored. The order in which the VMs are processed depends on the VM Recovery Options defined while configuring the plan.
If datastores added to the recovery location as Storage resources breach the capacity threshold before all the VMs are restored, the failure of this step for a VM from a critical VM group halts the plan. To troubleshoot the issue, configure the location to add more datastores and try running the halted plan again. To learn how to work with HALTED restore plans, see Managing Halted Plans.