Originally bearing the name Networkers, Cisco’s premiere network conference, CiscoLive, has connected, educated, and entertained attendees with a variety of backgrounds and skill sets since 1989. As the decades pushed on and technology followed suit, CiscoLive’s attendance grew; in 2019 alone, Cisco’s San Diego conference had gathered roughly 28,000 attendees—with an accompanying 74,000 mobile devices, ranging from cell phones to tablets. As a networks conference, Cisco, and tools such as InfluxData’s InfluxDB and Grafana, did not disappoint in its upholding of efficient network architecture processes to accommodate those devices, bringing in 650 network switches and over 2,300 wireless access points. Through this incorporation of a series of visualizations and network support facilities, Cisco demonstrated its prowess in top-tier network management—with the attendees as their audience.
DBTA recently held a webinar titled, “How Cisco Provides World-Class Technology Conference Experiences Using Automation, Programmability, Python, InfluxDB and Grafana,” featuring speaker Jason Davis, distinguished services engineer for DevNet at Cisco Systems, to discuss the precise approaches taken by Cisco during CiscoLive events that improve network performance, availability, resiliency, and observability through automation.
Davis ventured into critical tenants of Cisco’s network strategies which shape the success of sizable device environments. Among these is observability, which Davis highlighted as fundamental to getting the most out of your network. If you can’t observe and analyze trends ongoing within your network, how will you best leverage it? The addition of automation further necessitates observability, allowing you to be constantly up-to-date with your network’s processes and automated functions.
Metrics backed by time-series database InfluxDB, Davis offered, are the backbone to properly managed automated networks. Collecting data such as the number of wireless clients per wireless AP, signal strength, temperature, power consumption, availability, latency, onboard/offboard count, CPU/Memory/Interface/Errors on routers, switches, APs, servers, and storage—and the list goes on—offers useable insights that are then directly shared with conference attendees via uniquely individualized dashboards with Grafana.
The large device environment generated by CiscoLive requires a build-out from scratch, Davis stated, to fully demonstrate the efficacy of their network management methods. To support the needs of CiscoLive, Cisco implements a mobile containerized data center—a layered shipping container with the routers, core switches, compute, and storage necessary to run the event. Although this is a substantial effort, it is necessary to ensure that Cisco employees do not have to configure each network device, or physically adjust access points to illustrate their network management strategy. In the end, Cisco creates a modular, resilient, highly available, and highly performant network that uses layered security to provide accelerated access to whatever you require. Davis explained that network availability is similarly critical as observability, resiliency, and performance; it’s not only routers and switches that need network reliability, but 300-plus virtual machines—such as any digital signage, printers, and point-of-sale terminals.
Dashboards, underpinned by Grafana and InfluxDB, employ Cisco’s network metrics to visualize its performance for CiscoLive attendees to glean value, “warts and all.” Not only do these dashboards provide specialized views of time-series data for Cisco’s support team, it is accessible for attendees at all skill sets. Even if an attendee is not a network subject matter expert, whether they are programmers, cloud native workers, or simply interested in network tech with no prior knowledge, Cisco offers a dashboard for an extended audience that provides metrics in a way that resonates with them.
Davis further explained their strategies for network management as an accessible venture, as well as effective; with open-source tool InfluxDB, engineers like Davis spend more time on finding the instrumentation and telemetry that is meaningful to them, rather than figuring out how to store and retrieve it. You don’t have to be a classically-trained database administrator to utilize this method, said Davis—little database knowledge is required, allowing you to focus on your technological specialty. Employing open source tools is particularly beneficial, as access to continued product enhancements and flexible APIs support CiscoLive’s network infrastructure.
Successful network environments prioritize innovative technologies built to withstand substantial device landscapes, and Davis emphasized, are accessible to others. CiscoLive’s thorough network management methods are ones they share through visualizations they expose. Sharing your dashboards invites a malleable and informed network atmosphere that is not only available, but effective.
You can view an archived version of this webinar here.