Home > Articles > Advanced Features

Advanced Features

Chapter Description

In this sample chapter from Implementing Cisco HyperFlex Solutions, you will review the HyperFlex HX Data Platform disaster recovery feature and configuration steps needed to enable replication between two HyperFlex clusters.

From the Book

Implementing Cisco HyperFlex Solutions

Implementing Cisco HyperFlex Solutions

$47.99 (Save 20%)

Replication Networking Overview

The HyperFlex clusters that will replicate must have TCP/IP connectivity between them, and additional IP addresses must be provided to an internal IP address pool that the HX SCVMs will use. The minimum number of IP addresses required is the number of nodes in the cluster plus 1 additional address. More addresses than are currently needed can be placed into the pool to allow for future growth of the HX cluster. An existing VLAN ID and subnet can be used, although it is more typical to configure a specific VLAN and subnet to carry replication traffic that will traverse the campus or WAN links between the two clusters. The VLANs that will be used for replication traffic must already be trunked to the Cisco UCS fabric interconnects from the Northbound network by the upstream switches, and this configuration step must be done manually prior to beginning the HyperFlex Connect configuration. The bandwidth usage of the replication traffic can be set to a limit to avoid saturation of the interconnecting network links, or it may be left unlimited, in which case the maximum network bandwidth is set as the total available to the network. The bandwidth consumption will be directly affected by the number of VMs being protected and the frequency of their replication. The set bandwidth limit must not exceed the physical bandwidth, and it must be the same on both sites of the disaster recovery environment. The allowed low bandwidth is 10 Mbps, and the maximum latency supported with 10 Mbps is 75 ms.

The interconnection between the two clusters at the two sites can be done in several ways. In most cases, the uplinks from the HX clusters carry all the needed VLAN IDs on the same set of interfaces, including HX management, vMotion, storage traffic, guest VM traffic, and the replication traffic. In some cases, it is desired that the replication traffic traverse a set of independent uplinks, which is referred to as a split L2 topology. However, due to a technical limitation, the configuration of replication networking cannot accommodate a split L2 configuration. Specifically, a single Cisco UCS vNIC cannot carry multiple VLANs that traverse multiple uplink groups. Because the default configuration uses vmnic0 and vmnic1 to carry HX management traffic and replication traffic, both of those VLANs must arrive to UCS across a single set of uplinks. The replication subnets and VLANs used in the two sites can be different routed subnets, or they can be a single subnet if the WAN is using other technologies, such as Overlay Transport Virtualization (OTV). Replication traffic originates and terminates on the SCVMs running on each HX host.

Figure 6-2 shows the network topology of a two-HX cluster.


Figure 6-2 Replication Network

Replication Network Considerations

A replication network must be established between clusters that are expected to use replication for data protection. This replication network is created to isolate inter-cluster replication traffic from other traffic within each cluster and site.

The following is a list of considerations when configuring replication networks and pairing:

  • To support efficient replication, all M nodes of Cluster A have to communicate with all N nodes of Cluster B, as illustrated in Figure 6-3 in the next section.

  • To enable replication traffic between clusters to cross the site boundary and traverse the Internet, each node on Cluster A should be able to communicate with each node on Cluster B across the site boundary and the Internet.

  • The replication traffic must be isolated from other traffic within the cluster and the data center.

  • Creating this isolated replication network for intercluster traffic involves creating a replication network on each cluster and then pairing clusters to associate the clusters and establish M*N connectivity between the clusters.

  • IP addresses, subnet, VLAN, and gateway are associated with each replication network of each cluster. You must configure the corporate firewall and routers on both sites to allow communication between the clusters and the sites on TCP ports 9338, 3049, 9098, 4049, and 4059.


Figure 6-3 M*N Cluster Connectivity

M*N Connectivity Between Clusters

Figure 6-3 shows the connectivity between two clusters: a primary cluster with four nodes and a secondary cluster with four nodes.

Configuring a Replication Network in HX Connect

A replication network can be configured only once. Once a replication network is configured, you can edit the available IP addresses and the networking bandwidth. Here’s how you configure the network:

  • Step 1. Log in to HX Connect as administrator by entering the username and password, as shown in Figure 6-4, and clicking Login.

    FIGURE 6-4

    Figure 6-4 HX Connect Login Page

  • Step 2. In the HyperFlex Connect screen that appears, select the Replication page and then click Replication Configuration > Configure Network (see Figure 6-5).

    FIGURE 6-5

    Figure 6-5 Replication Network Configuration

  • Step 3. In the Configure Replication Network dialog box, on the VLAN Configuration tab, enter the network information:

    • Select an existing VLAN: Click this radio button to add an existing VLAN(see Figure 6-6).

    • Create a new VLAN: If you select this radio button, the following fields become available (see Figure 6-7):

      • VLAN ID: A VLAN that is different from the HX Data Platform management traffic network and data traffic network

      • VLAN Name: The name of the VLAN

    FIGURE 6-6

    Figure 6-6 Existing VLAN Assignment for Replication Network

    FIGURE 6-7

    Figure 6-7 VLAN Creation for Replication Network

    • UCS Manager host IP or FQDN: The host IP address or FQDN for the UCS Manager

    • Username: The administrator username for Cisco UCS Manager

    • Password: The administrator password for Cisco UCS Manager

  • Step 4. Click Next.

  • Step 5. In the IP & Bandwidth Configuration tab, set the network parameters and the replication bandwidth (see Figure 6-8):

    • Subnet: The subnet for the replication network

    • Gateway: The gateway for the replication network

    • IP Range: The range of IP addresses for each converged node in the cluster plus one additional IP address (that is, N+1)

    • Add IP Range: Button for adding the specified IP range

    • Set replication bandwidth limit: The bandwidth that the replication network is allowed to consume

    • Set non-default MTU: MTU other than the default, which is 1500.

    FIGURE 6-8

    Figure 6-8 IP and Bandwidth Replication Settings

  • Step 6. Click Configure.

  • Step 7. In the Test Configuration tab, shown in Figure 6-9, check the replication network configuration.

    FIGURE 6-9

    Figure 6-9 Replication Network Test Configuration

4. Replication Pair Overview | Next Section Previous Section

There are currently no related articles. Please check back later.