The SVM disaster recovery workflow includes creating and preparing the destination SVM, activating the destination SVM in the event of a disaster, reactivating and reconfiguring the source SVM for disaster recovery.
The source and destination clusters must be peered.
This task provides a high-level overview of the tasks involved in setting up disaster recovery for an SVM.
Express Guides are available that provide detailed information about configuring SVM disaster recovery.
The following illustration shows the source and destination SVMs in the preparation phase:
After the preparation is complete, the destination SVM is in the stopped state.
The data flows from the source SVM to the destination SVM, and the flow is unidirectional. SnapMirror transfers occur based on the SnapMirror policy schedule. You can also perform SnapMirror updates whenever required.
If the source cluster reboots, then the source SVM is operationally stopped and is locked for any management operations to avoid data corruption in case data is accessed inadvertently from both the source and destination SVMs.
The following illustration shows the impact of a disaster on the source SVM and shows that the destination SVM is activated:
In this phase, the SnapMirror relationship is in the Broken-off state.
When the source cluster and SVM are completely destroyed, the cluster administrator of the source cluster creates the cluster and source SVM. The cluster administrator resynchronizes the data and configuration from the destination SVM, activates the source SVM, and protects the new source SVM.
If the source SVM is available, the cluster administrator resynchronizes the data and configuration from the destination SVM to the existing source SVM, activates the source SVM, and protects the source SVM.
The following illustration shows the configuration and data flow during the reactivation phase of the source SVM:
The data flows from the destination SVM to the source SVM.