The virtual desktop infrastructure for, most of the remote display agreement will take part in bandwidth, within the scope of available bandwidth, adjust the bandwidth to ensure user experience. End users can experience to improve network performance, because their virtual desktop might periodically to gigabit speed of Internet connection to the server.
One example is, the user from remote desktop access to Microsoft Outlook, and enterprise Microsoft Exchange server using the same network. WAN I/O affects the user experience, if multiple user session at the same time produce, the effect will be worse.
Not any one enterprise can fully control the network layer. VDI network bottleneck may be the cause of the user in a can of coffee shop on the Internet by WiFi connection virtual desktop. One day in the different period of time, the user may use high-speed WiFi connection to their virtual desktop, but then they will be forced to use equipment not reliable 3 g network.
Allows the user to use 3 g connection virtual desktop, cost is expensive, but also can increase the risk of VDI network bottleneck. Because the remote display cost bandwidth and monthly flow. Enterprise check the equipment use, and as a service data plan choice basis.
The remote display not only restricted to VDI bandwidth factors. The client to virtual desktop connection delay and bad packets are able to affect stability and usability.
With the increase of the delay, the network data through all previous number will increase. Therefore, even if the source network and target network resources are abundant, the enterprise can control quality or management between the client and the server of the Internet infrastructure. This may be the reason for the existence of the variability, especially in the United States and northern Europe.
The user experience delays, still able to connect to the desktop, but the keyboard will delay, the screen update will slow. Generally speaking, both ends of the delay to 220 milliseconds (msec), can cause slight VDI network performance problems. The ideal delay is less than 150 msec. Any less than 250 msec delay, through the flow plastic and WAN optimization can deliver a certain performance to ease congestion peak.