In a previous article we have underlined the importance of videoconferencing systems and their pervasiveness in corporate life. From a technical point of view, efficiently supporting these tools on a corporate infrastructure, as well as the related network flows, is a task that creates interesting, even if sometimes complex scenarios. One of the aspects that often fails to be perceived is that the choice of platform is fundamental, the different products can offer radically different operating modes (in architectural terms) and condition the infrastructure configuration choices.

All the traffic generated by users’ simultaneous sessions falls on corporate infrastructures, which support the work of tens, hundreds or thousands of workers. But that’s not all. In most cases these infrastructures are conveniently managed through centralized resources (in on-premise servers or in the cloud) which are allocated and shared for the different users in order to optimize the overall IT management of the organization, for example through the use of desktop virtualization technologies (VDI). In these cases, the computational burden of encoding and decoding each audio-video stream requested by each user would naturally fall on the related virtual machine and, consequently, the overall burden of all streams would be concentrated on the same company server (or farm) in charge (terminal node also of the related traffic generated by the application). This therefore entails an extremely high potential overall consumption of resources when the use of collaboration tools in the company is frequent and pervasive.

For this reason, the main technologies that have been present for years in the world of virtualization have made available customized solutions for the main video conferencing applications to optimize their management on large VDI infrastructures. These optimizations allow “offloading” and “peer to peer” of audio and video streams after starting the call. Concretely, the computational load is “downloaded” from the server which is exempt from it, delegating the burden to local devices. Once the call has been made, these are put into communication directly with each other as peers, mediated by the global collaboration application server. Finally, the optimization tool composes the video stream relating to the virtual desktop directly “on the end user’s screen”, filling it with what is related to the video conferencing application area, without having to pass through the virtual machine server.

For example, both Citrix and VMware have made available optimization packs to handle leading video conferencing solutions on the market, such as Zoom, Microsoft Teams and Cisco WebEx.

These optimizations simply require the client-side installation (user workstation) of plugins/add-ons that have been made available for both endpoints with Windows and Linux-based operating systems. Praim supplies these plugins already integrated on its hardware and firmware, both with Windows 10 IoT or ThinOX operating system, or the possibility of distributing them on all the workstations through the ThinMan management console.

In addition to the application of these plugins, some additional precautions are also required. Indeed, some of these components require specific compatibility between the plugin version and the version of the collaboration application (server) in use. Other configurations of the VDI part that the various companies describe well on their documentation pages must also be managed (depending on the case, some types of redirection may need to be disabled, virtual channels activated, etc.). Finally, it should be remembered that these optimizations directly connect the peers through the Internet connection, therefore the configuration of corporate firewalls will also have to take account of the changes, allowing the endpoints to directly reach the application server on the Internet through specific ports to be configured.

As far as web-based collaboration tools are concerned, there’s an interesting possibility, thanks to the feature called Browser Content Redirection. This allows certain web addresses specified in the ACL policies to be reproduced on the endpoint, (for example from Citrix). The local browser is overlaid on the remote session browser, transparently to the user. In this case, some attention must be paid to the ACLs of the websites to be redirected, especially as regards the authentication session.