VDI network best practices: Ensure elasticity and Redundancy

When a company migrates from a traditional PC to a virtual desktop, the VDI network traffic increases significantly because the application and user data need to be transmitted through LAN and WAN connections to remote users.

To plan to support a large number of virtual desktop network architectures, We need to select the appropriate display protocol, increase a considerable amount of bandwidth, and set VDI network access policies to ensure security.

VDI uses stateful access between end-to-end client devices and backend architectures, such as accessing the backend from a PC, tablet, or smartphone. Stateful means that the client is always connected to the backend virtual desktop environment. Essentially, this connection is similar to a VoIP (Voice over Internet Protocol) application.

An unstable network may cause data packet loss, which greatly affects the user experience. Taking VoIP as an example, the phone may hear noises or even disconnect. In a virtual desktop, the mouse and keyboard are delayed and unresponsive.

A good VDI network should provide continuous and stable data streams. When administrators view the network, they often pay attention to the bare throughput. In fact, the key factor is stable and reliable networks without packet loss. During VDI deployment, the voice network should have the necessary fault tolerance capabilities, including Nic pairing, multiple switches, redundant links to the backbone network server or data center.

These components may be a strong point in the architecture, but do not take it lightly. You also need to ensure that the architecture is still available when a nic, physical switch, or server site fails. However, few enterprises actually test the disaster tolerance capability of the physical architecture.

In addition, powerful monitoring tools are needed to identify whether N + 1 Redundancy works properly. The elastic design of some networks is seamless. When component A fails, the user will not notice it until component B also fails. Establishes a review mechanism for monitoring tools to ensure that alarms are generated immediately when a component fails.

Once the enterprise can ensure the redundancy of the physical architecture and the normal operation of the monitoring mechanism, the next step needs to focus on the weaknesses in the architecture-typically the WAN environment. Especially when enterprises need to provide services for remote users working at home, including users working in the office through VDI networks, it is more critical.

If you imagine that most users are sitting in the office and using the same computer like dumbert every day, that would be wrong. The user and management experience has completely changed. Even if you are not on a regular business trip, you may want to work on weekends or sometimes at home. They do not want to always sit in the office while working.

At least, you should establish two independent channels to connect to the VDI environment location. It is best to have two different service providers. Ideally, their architecture is completely independent.

In the WAN field, the architecture of a supplier acting on behalf of other companies is also very common, so on the surface of a completely independent connection may be affected by the downtime of the same upstream supplier, this affects the entire architecture. Before signing the contract, it is best to sit down with the network service supplier to discuss the establishment of a suitable SLA level, which can ensure the provision of clear service quality. SLA can contain certain penalty conditions, assuming that the supplier cannot provide the corresponding service according to the negotiated standard.