You are asked to deploy a 2-node ONTAP Select cluster. The physical hardware is located in two different locations and you want the aggregates mirrored synchronously between the two data centers. You must ensure high availability and automated switchover.
In this scenario, which two steps would you take to fulfill this requirement? (Choose two.)
To deploy a 2-node ONTAP Select cluster with physical hardware in two different locations and ensure high availability with automated switchover, you need to use the ONTAP Mediator service to maintain quorum between the nodes. This is essential for managing failovers and preventing split-brain scenarios when a node fails or loses communication. Additionally, the ONTAP Select Deploy virtual machine, which includes the mediator service, should reside in a third location separate from the two data centers to ensure proper monitoring and failure management.
I would say A and C https://docs.netapp.com/us-en/ontap-select/reference_plan_ots_two_node.html When a cluster consists of two nodes, it is not possible to achieve the quorum required when a node fails or loses communication. To resolve these types of split-brain situations, every instance of the ONTAP Select Deploy utility includes a mediator service. If you have one or more active two-node clusters, the ONTAP Select Deploy virtual machine administering the clusters must be running at all times.
Agreed. As with all two-node ONTAP Select clusters, there is a separate mediator service contained in the Deploy virtual machine that monitors the nodes and assists in managing failures. With the enhanced distance available with MetroCluster SDS, this creates three distinct sites in the network topology. Latency on the link between the mediator and a node should be 125 ms round-trip or less.
B&C. You use the ONTAP Select Deploy to install the nodes on the physical hardware and the ONTAP Select Deploy VM has the mediator built in of which resides at a third location other than the two data centers
This would make sense as A is already covered by default once you deploy the Ontap Select VM
This is wrong. ONTAP select installs into either KVM or ESX so B is wrong
Select to be installed on hypervisor not on HW, so B is wrong
When a cluster consists of two nodes, it is not possible to achieve the quorum required when a node fails or loses communication. To resolve these types of split-brain situations, every instance of the ONTAP Select Deploy utility includes a mediator service https://docs.netapp.com/us-en/ontap-select/reference_plan_ots_two_node.html B is definitely wrong, nodes can't be installed on the physical hardware.
The scenario is for an ONTAP Select MetroCluster SDS deployment. https://docs.netapp.com/us-en/ontap-select/reference_plan_ots_mcc_sds.html https://docs.netapp.com/us-en/ontap-select/concept_usecase_mc_sds.html Not A, "As with all two-node ONTAP Select clusters, there is a separate mediator service contained in the Deploy virtual machine that monitors the nodes and assists in managing failures." Not D, SnapMirror licensing isn't needed for an ONTAP Select MetroCluster SDS deployment.
To reference why not D, https://docs.netapp.com/us-en/ontap-select/reference_lic_ontap_features.html#ontap-features-automatically-enabled-by-default
A & D. Use ONTAP Mediator service to ensure that there is a quorum between the two nodes. Install the SnapMirror Sync license on the cluster to enable synchronous replication of data between the two nodes. How can the mediator service check for quorum if quorum is not possible? Well, he Mediator service in an ONTAP cluster can check for quorum even if quorum is not possible through the use of a witness node. A witness node is a third-party storage device that acts as an independent and neutral third party that can provide a vote to help determine quorum.
How can the mediator service check for quorum if quorum is not possible? I would consider A a trick answer. mediator is required to properly identify the active node during a failover and avoid any split-brain scenario in which both nodes remain independently active during a network partition, but it is does not check for quorum as quorum is not possible.