A FlexPod converged infrastructure solution includes NetApp storage, Cisco compute and networking components, and the application stack running on the infrastructure. FlexPod solutions are highly available, highly scalable, and highly flexible by design and they are adopted by companies world-wide for business-critical workloads and applications.
Depend on your business requirements, there are several disaster recovery and business continuity approaches that you can consider as illustrated in Figure 1, using traditional approaches and NetApp ONTAP SnapMirror capabilities, to achieve different levels of data protection.
Figure 1) FlexPod disaster recovery and business continuity flexibility.
Traditionally, backup applications, such as Veritas NetBackup and products from Veeam and Commvault, are used to perform scheduled backup and on-demand restore for data protection and recovery. With ONTAP, you can take advantage of its flag-ship SnapMirror functionality to replicate data to targets on-premises or in the public cloud.
Both SnapMirror Business Continuity and MetroCluster solutions provide zero recovery point objective (RPO) with no data loss and very low recovery time objective (RTO) to meet your requirements. SnapMirror Business Continuity supports SAN volume replication to another ONTAP system which can be a different hardware platform and you choose a number of data volumes for data protection with replication. On the other hand, MetroCluster solution protects all your workload data, both SAN and NAS, and requires the two ONTAP clusters that are paired together to have the same hardware platform. Implementing MetroCluster solution with FlexPod gives your business a special simplicity advantage. Once it is configured, all your business-critical data services will be available when faced with many single-point-of-failure scenarios or even a site disaster.
The following are important benefits the FlexPod MetroCluster IP solution provides:
The solution is highly flexible with choices of storage, network, and compute components to meet your requirements.
The solution is highly reliable with tier-1 enterprise HW from NetApp and Cisco backed by a cooperative support model.
The solution can be scaled up and scaled out to meet your continued business growth.
The solution supports site distances up to 700Km with zero recovery point objective (RPO) and very low recovery time objective (RTO) for no data loss and quick recovery.
The solution requires one-time configuration, and it continues to monitor and protect all your SAN and NAS workloads and application data to ensure business continuity.
At a high level, a FlexPod MetroCluster IP solution consists of two FlexPods, located at two sites separated by some distance, that are connected and peered together to provide highly available data services for business continuity. The MetroCluster IP storage replication network can utilize a dedicated network or it can be integrated into the data center network using the compliant switches configuration, when the switches and network configuration meet the requirements.
For a small solution deployment that might be cost-sensitive, simply extending Layer 2 network across sites will require the least amount of hardware to implement a FlexPod MetroCluster IP solution. You can use the Cisco Nexus switches in the FlexPod as compliant MetroCluster IP switches if they meet the requirements, thus eliminating the need for dedicated MetroCluster IP switches. You can achieve additional cost savings by sharing the Inter-Switch Links (ISL) between sites for MetroCluster IP and non-MetroCluster IP usage if they have sufficient bandwidth as shown in Figure 2. A NetApp ONTAP Mediator monitoring the solution from a third site enables the solution to perform an automated, unplanned switchover when one of the sites experiences a failure.
Figure 2) FlexPod MetroCluster IP solution architecture with compliant MetroCluster IP switches. 
If you are designing a solution for a much larger multi-site network environment, utilizing VXLAN or Cisco ACI, the FlexPod MetroCluster IP solution architecture in Figure 3 below with dedicated MetroCluster IP switches might work better when scaling out the solution for MetroCluster IP and non-MetroCluster IP traffics separately to meet your network requirements.
Figure 3) FlexPod MetroCluster IP solution architecture with dedicated MetroCluster IP switches. 
NetApp collaborated with Cisco and published a white paper on FlexPod MetroCluster IP solution recently. In this solution, the VXLAN EVPN Multi-Site network architecture is used to provide business continuity / disaster recovery by interconnecting data center fabrics in two sites (Figure 4). The Cisco VXLAN fabric is managed by Cisco Data Center Network Manager (DCNM) deployed outside the fabric. It is deployed as a cluster of multiple nodes for high availability. The network fabric consisting of spines and leaf switches is built using Nexus 9000 series switches capable of supporting VXLAN overlays. The fabric provides seamless Layer 2 extension and Layer 3 forwarding for both infrastructure and application and services VMs hosted on the infrastructure.
Figure 4) Cisco DCNM simplifies the FlexPod MetroCluster IP solution VXLAN EVPN multi-site fabric management. 
For this solution, dedicated MetroCluster IP switches are used to provide the storage fabric for MetroCluster IP and ONTAP intra-cluster communications within each site and between the two sites. (Figure 5) For the A700 storage controllers, each A700 node is connected to both Cisco Nexus 3132Q-V 40G switches at its local site. The connections are used for intra-cluster and MetroCluster IP traffic from the intra-cluster and MetroCluster IP node ports. The intra-cluster ISL between switches at each site is required for the ONTAP cluster communications. The MetroCluster IP ISLs between sites carry traffic for MetroCluster IP storage data and nonvolatile RAM (NVRAM) replication data between sites. The redundant switches and redundant connections configuration in MetroCluster IP provide a highly available multi-site storage solution.
Figure 5) Dedicated MetroCluster IP switches with cross-site connectivity provides network fabric for data replication. 
All FlexPod validated solutions undergo thorough testing and this FlexPod MetroCluster IP solution is no exception. The FlexPod Datacenter with Cisco VXLAN, NetApp MetroCluster IP, and VMware vSphere 7.0U1 solution is validated for successful infrastructure configuration, high availability, and business continuity across two sites. A variety of test cases are used to verify solution functions and to simulate partial and complete site failure scenarios. For the VMware cluster deployed on the infrastructure, the VMware High Availability (HA) and Distributed Resource Scheduler (DRS) features are enabled so the virtual machines can be restarted on the available hosts when a site disaster occurs.
FlexPod converged infrastructure is highly available, scalable, and flexible and you can start small and grow your solution infrastructure as your business grows. The FlexPod MetroCluster IP Datacenter uses an active-active data center design to provide business continuity and disaster recovery. The solution interconnects two data centers deployed in separate, geographically dispersed locations. The solution synchronously replicates data between sites to protect business-critical data services against site failure to achieve zero recovery point and low recovery time objectives. The NetApp ONTAP Mediator and VMware vCenter deployed at a third site help you monitor the MetroCluster IP operations and manage the stretched VMware cluster solution. To learn more about how the solution provides resiliency and protection against many single-point-of-failure scenarios as well as a site failure to ensure business continuity, please check out the white paper and the additional resources listed below.