Hyper Converged Infrastructure (HCI/aSV)

Sangfor HCI and aSV provide a unified infrastructure combining compute, storage, networking, and built-in security to simplify deployment, operations, and services.
{{ $t('productDocDetail.guideClickSwitch') }}
{{ $t('productDocDetail.know') }}
{{ $t('productDocDetail.dontRemind') }}
6.11.1R1
Hyper Converged Infrastructure (HCI/aSV) {{ breadTitle }} User Manual HCI Cluster Initialization Four Networks Multiplexing (Optional)
{{sendMatomoQuery("Hyper Converged Infrastructure (HCI/aSV)","Four Networks Multiplexing (Optional)")}}

Four Networks Multiplexing (Optional)

{{ $t('productDocDetail.updateTime') }}: 2026-01-05

Description

When the physical NIC of the server is insufficient, the four networks multiplexing can be adopted. The four network multiplexing is a network planning scheme in which the management network, physical edge network, overlay network, and storage network interface reuse one aggregation interface. The scheme provides NIC-level redundancy protection, and the failure of a single NIC will not affect the service.

Precautions

  1. A single interface does not support the integration of four networks. It can support the aggregation of four physical interfaces for integrating four networks at most.
  1. The server NIC aggregation supports the active/standby mode (default) and five load balancing modes according to IP, MAC address, interface polling, and four-layer information. Using the IP mode for load balancing is recommended, and the active/standby mode is not recommended. The access layer switch interface needs to be configured with the corresponding interface aggregation mode. The aggregation peer switch in the active-standby mode can be configured without aggregation.
  2. The current aggregation mode does not support the direct aggregation of two existing role network interfaces. You need to move the role to an idle network interface before aggregation.
  3. It is recommended to set the upper limit of traffic when the four networks of management, VXLAN, service, and storage are integrated to ensure the stability of network traffic. After the speed limit is enabled, you can view the throughput of the management interface, VXLAN interface, Edge interface, and storage interface on the Nodes > Summary page.
  4. In HCI6.8.0, the following network cards are supported for the four network multiplexing function: Mellanox CX4, Mellanox CX6, Intel X710 and Intel X722.
  5. It is recommended to configure VLAN for network isolation in each network plane. VLAN sub-interfaces are configured for management, VXLAN, and storage. VLAN sub interfaces are not supported for service interfaces. Service VLAN can only be configured through the interface group of the virtual network.
  6. After the active upgrade from the old version, the functional operation will fail if the aggregation interface already exists in the original environment, but it will not affect normal use. The operation can be supported after the node is restarted.
  7. The device upgraded from the old version enables the integration of four networks. The Intel network card will interrupt the network for 30 seconds (restart DP + switch STP convergence). The Mellanox NIC needs to restart the physical node before it takes effect.
  8. The motherboard must support VT-D (supported by all CPUs after the third generation). The new delivery aServer will turn on when installing the system. If the vt-D function is not enabled before upgrading the old version or the third-party device, it needs to be restarted to take effect.

Prerequisites

  1. The multiplex interface of interface sharing must be at least two 10 Gigabit interfaces.
  1. VLAN is recommended to be configured for network isolation in each network plane, so the switch interface connecting the HCI host must be configured as a trunk in advance. And configure the planned VLAN in the relevant switch.

Steps

  1. Click Nodes > Physical Interfaces on the console.

  1. Since it does not support the direct aggregation of two existing role interfaces, you must first multiplex the Overlay Network Interface to the management interface. Then, Navigate to Communication Interfaces > Overlay Network Interface > Settings to multiplex the Overlay Network Interface to the management interface.

  1. Select OK.

  1. After all node interfaces are adjusted, click OK. Check the Enable high performance mode checkbox (this operation will set the MTU of the network port to 1600. You need to enable jumbo frame support on the physical switch. Otherwise, the network will be blocked).

  1. The Overlay Network Interface has been adjusted to the management interface. The original Overlay Network Interface has no role and can be aggregated.
  2. Aggregate the planned network interfaces, and enable the storage reuse function when creating the aggregation interface. Aggregation mode is recommended to load according to MAC or IP address.
  3. Click Add Multiple Aggregate Interfaces, and select the planned interface. Next, select the aggregation network interface working mode. The active-standby mode is not recommended, and the network interface utilization is poor (after configuring the non-active-standby aggregation mode, you need to adjust the aggregation mode of the connected switch to match the platform. Otherwise, you will not be able to access the platform). Check the Storage Network Interface Multiplexing checkbox, and click OK.

  1. Confirm to enable Storage Communication Interface Multiplexing. Before enabling the Storage Communication Interface Multiplexing function, you need to restart the network service, temporarily interrupting the physical node and the VM networks running on that specific node for several minutes. Please operate with caution.
  2. Click OK to bulk add the aggregation interface of the node. Please ensure that the IP address and VLAN ID are configured correctly to avoid network interruption after the configuration takes effect.
  3. As shown in the figure below, click Reset Storage Area Network, and adjust the storage communication interface.

  1. Select Shared Mode. Click Next, select the aggregation interface with the previously enabled Enable storage network interface sharing. Fill in the planned VLAN ID and interface IP address, and click OK.
  4. Click Edit Configuration and adjust the Overlay Network Interface interface.

  1. Click IP Address Pool.

  1. Click Create for the new pool and fill in the planned IP address/netmask.

  1. Under Physical Interfaces, select Add Multiple VLAN Subinterfaces.

  1. Change the overlay network interface to the newly created VLAN subinterface. It will cause a short network interruption.
  2. Navigate to Nodes > Communication Interfaces > Edge-Connected Interface. Then, click Add New Edge, select the available network interface for all nodes and click OK to create the Edge.

Graphical user interface, application Description automatically generated

  1. Select the Edge and click Connect To Node.

  1. Click New to select the network interface and node.

  1. The configuration for the physical edge network interface is done. The physical edge interface does not support configuring VLAN sub-interfaces. The edge interface VLAN can only be configured through the interface group of the virtual network.
  2. Select the sub-interface of the management network interface and configure the planned VLAN ID for it.
  3. All interfaces of interface sharing have been configured.
  4. It is recommended to set the upper limit of bandwidth/traffic in the case of management, VXLAN, service, and storage interface multiplexing to ensure the stability of network traffic. When creating a cluster, it is recommended to configure 3Gbps for the management interface. The service interface and VXLAN are configured with 5Gbps. The storage interface is configured with 8gbps. Subsequently, the traffic of each network plane is observed and then optimized and adjusted.