Mar 16, 2020

Understanding Network Impacts of Increased Online Learning

While COVID-19 is creating unprecedented uncertainty, Research and Education Networks such as CENIC's CalREN will play an important role in supporting our educational institutions as they increasingly move to online learning. Network usage patterns may vary greatly from traditional face-to-face learning environments, in part because of the massive shift of users from purpose-built campus networks to commercial, residential network services.

In an effort to contain the spread of COVID-19, a number of California’s universities, colleges, and schools have already suspended in-person classes and turned to use of online learning. As remote learning increases, one critical consideration will be users varying levels of broadband access at home locations and whether users have any access at all. At a recent meeting, CENIC member institutions suggested colleges and libraries might augment and extend their public WiFi capabilities to help bridge this gap. One idea of offering drive-up wireless hotspots to the community might be a significant contribution to K-12 schools, colleges, and the general public weathering this situation.

Another pressing issue as online learning increases will be the capacity for conferencing tools, such as Zoom, and Learning Management Systems, such as Canvas, to accommodate a massive influx of users. While CENIC’s network has ample capacity, should the application become overwhelmed, several CENIC member institutions have suggested that backup tools such as Google Hangouts Meet, Office 365, Skype, GoToMeeting, and BlueJeans could be used.

As we monitor this rapidly evolving situation, providing continued and uninterrupted service to our members while maintaining the safety and well-being of our employees is our top priority. CENIC has prepared this information as a resource for members who have moved to, or are contemplating moving to, online instruction.

How Does Network Access Change with Online Learning?

When students and faculty are on campus, well-tuned campus infrastructure is blended with remote cloud computing resources through R&E networks like CENIC’s CalREN and Internet2. R&E networks are provisioned with substantial capacity for research data and also afford “headroom” for activities like Learning Management Systems and video-based online learning services that are increasingly hosted off the campus in cloud computing environments. Network traffic patterns are likely to change as more universities, colleges, and schools implement COVID-19 “work from home” and “learn from home” models. This shift in the location of users will have an impact on the way in which online resources are reached.

It is important to note that the following scenarios depict typical traffic patterns. There will be cases where a different traffic pattern will occur or specialized engineering is in place that may impact specific campus patterns.

Scenario 1: Connectivity to Online Resources from CalREN-Connected Sites

There are two ways of connecting to CalREN: dark fiber and lit circuit. With dark fiber, when students and teachers are located at a CalREN-connected site, they reach cloud-based services by the most direct route: from their local network to CalREN and then to the service. If the service is hosted on AWS or another cloud provider, they then connect either directly from CalREN to the cloud service provider, or traverse CalREN, connecting to the provider (see white lines in Diagram). With a lit circuit, which is used by the majority of our non-university members, there is the added complexity of a commercial service provider managing the circuit between the site and CalREN. (see green lines in Scenario 1).

Scenario 2: Connectivity to Online Resources from Home

When students and teachers move to home locations, they typically connect via a commercial residential ISP. In this case, the network traffic from home to the cloud providers that host instructional resources, such as learning management systems, video applications, and collaboration tools, will almost exclusively use commercial networks and will no longer traverse R&E networks (see yellow lines in Scenario 2). While the exact configurations of these commercial ISP networks may be opaque, they are generally engineered with less “headroom.” Some congestion of these residential networks and their interconnections to cloud providers may exist from home networks that could impede the performance of online learning in unanticipated ways. Operators may wish to consider contingencies for poor-performing online experiences if/when these issues are diagnosed.

Scenario 3: Access to Campus from Home Using a VPN to Gain Access to Online Resources in the Cloud or Other Resources on the Campus

Another scenario is for faculty and staff to use a virtual private network (VPN) to establish a secure connection to the campus. This allows those users access to campus resources (e.g. ERP systems used by administrators), or to use the campus connectivity to reach cloud-based online services. In this scenario, the VPN will tunnel network traffic through commercial ISP(s) to reach the campus network (see the blue lines in Scenario 3). Access to resources, whether local to the campus or cloud-based, will work as if the faculty or student were physically present on the campus. Operators should consider the potential increased load this may place on their VPN appliance(s) and factor in additional user license requirements.

What Are the Broadband Network Issues to Consider in Implementing Online Learning?

Varying levels of access to broadband from home is an important consideration. Students and faculty who are off-campus may have sufficient connectivity at home from their local ISP or may be able to upgrade to a higher bandwidth connection. For those without broadband at home, options include identification of publicly accessible locations with good broadband connectivity, such as public libraries connected to CalREN, to the degree that these are open and available, or other available WiFi hotspots.

How Is CENIC Managing the CalREN Network to Help with Online Learning?

While we have significant unused capacity to the commodity Internet, we are working with our providers to increase that capacity to ensure all of our members who provide resources to their students, staff, and other stakeholders experience no bottlenecks to/from the CENIC Network. There are, however, potential impediments on the commercial Internet side, which are dependent upon how people are connected at home — if they have a broadband connection at all — who connects them, and how responsive these ISPs are to the current changing social environment and the demands this puts on their networks. Commercial networks will also face supply-chain issues, as all networks have been, where components are manufactured abroad.

CENIC will also continue to leverage our robust peering and caching infrastructure to provide the highest quality of service possible, especially as our member institutions rely more heavily on third-party, cloud-based services for their students and staff. While we already have significant connectivity to major cloud service providers, we will continue to monitor the situation and make adjustments/additions as needed.

Many thanks to Internet2 for providing its national model of online learning impacts on networks, which CENIC was able to adapt for its regional audience to create this document.

Related blog posts

Network Traffic Analysis Shows Changing Activity Patterns During COVID-19 Pandemic

Even as campuses, schools, libraries, and cultural institutions have closed amid stay-at-home orders, CENIC’s networks have remained a vital part of the Internet ecosystem for online learning and community engagement, remote access to research data and specialized computing facilities, academic medicine and clinical care, and work-from-home operations.

CENIC and Alameda Peer to Connect Students to Saildrone and Berkeley Lab

A new SFI agreement between CENIC and the City of Alameda is a model for local governments that wish to connect technology and R&D companies in their region with a state’s research and education institutions.