Hi, everyone, we're back in this modern campus network designed series. We hit trends and challenges, moved into campus architectures and covered some cool concepts related to cloud edge and hybrid solutions. Now we're ready to move on to NMS evolution. I want to talk about typical industry solutions, best practices and technologies, starting with issues related to disparate systems. In the first video of this series, I talked about classic network management systems, or NMS challenges. Separate systems in different regions of your deployment make it difficult to see national or international patterns and trends. Then you have these separate silos based on technology WLAN, LAN and WAN connectivity to remote offices and services. And this means that to troubleshoot a problem, you must often manually move information from one tool to another tool to discover the root cause of an issue. Suppose you have some host having issues on the WLAN with a particular Mac address, you need to pull data from the wireless system about APs, SSIDs and controllers. From the LAN system about switches, ports and status, and perhaps from the WAN system, about routers, paths and layer three security, then piece all that information together and puzzle out a solution. The same holds true for configuration management. You want to add a new WLAN. So you access the WLAN system to configure APs, SSIDs, VLAN mapping and WLAN security. Then access the LAN system to configure ports, VLANs layer two security and maybe quality of service or QoS and trunking. Then perhaps access the WAN system to accommodate this new WLAN with the appropriate sub netting routing and layer three security, this all might be in three different tools with three different approaches. Automation can be built, but you're still stitching together these disparate systems, it's frustrating. The industry is responding with a couple of key approaches. For example, vendors air starting to combine strategic elements into unified management tools, allowing prepackaged turnkey approaches to unification. And some organizations write completely custom tooling to pull together automation or information from multiple systems. And these solutions can get labor intensive. And they might not provide the tight integration you're looking for. But they help, and they can get one aspect of the equation right by allowing for multi-vendor environment, NMS evolution means that you must be able to integrate with many different solutions. I think a roof is approach is pretty cool. Unifying these disparate systems into a common platform with a common data lake wired wireless. And WAN across your entire deployment, this common big data pool can be analyzed with cloud based AI and ML systems to detect trends, see patterns and alert you of impending issues and advise you how to a handle those issues. And Aruba's edge services platform, or ESP, ties this all together with Aruba's central, giving you a single portal into all that information and intelligence. But NMS evolution isn't just happening in the cloud. It's also happening in the fabric of our campus networks in the switches, multi layer switches and controllers at the heart of your campus architectures. For many years, network admins have been dealing with the old simple network management protocol, or SNMP, which has evolved from version 1 to 2.3c. And on version three, which includes more robust security mechanisms and some efficiency advantages. Now it still works okay, but its poll driven approach to telemetry is a bit dated. It remains heavily used. But there are opportunities to find a more streamlined, universal approach to configuring, logging and diagnosing these systems. And the old command line interface, or CLI, with flat text configuration files, is getting a bit old fashioned as well. The old CLI is great for human to machine interaction, but not so great for machine to machine interaction. It just doesn't lend itself to the scalability and ease of management requirements of modern campus designs. But wouldn't it be nice to have a unified infrastructure with a single operational model across the infrastructure? I like that the Aruba CX switches are built on a micro services based operating system with an open rest API available for configuration and management functions. You get easy, secure cloud connectivity with modern HTTP-based telemetry that uses or optimized publisher subscriber model, as opposed to the cumbersome polling method of SNMP. This facilitates zero downtime upgrades and moving to a more centralized control and management plane in cloud or on premises. It also facilitates more automation related to typical network administration task and scalability. A single customer can manage up to 100,000 network devices and one million client devices. Check it out by zooming in on one of these switches. Inside, you have a CPU board and kernel with all management history and protocol and network analytics based around a state full database. Of course, you have high speed communications down to the application specific integrated circuits or ASIC hardware for network connectivity. Now here's the cool thing. You can access all this modern internal architecture, using the same command line and graphical interfaces you know and love, but are also ready to take advantage of that rest API for cloud like services, automation and more. Another aspect I love about these modern devices is the continued march toward a kind of optimized edge cloud delivery model. Networking used to be about simply moving the packets around. But as the infrastructure got more complex, tools were needed to optimize, upgrade and provide additional value to the business is leveraging them. Vendors have capabilities like this that come in many flavors. Some of these may include integrations to third party systems to automatically trigger things like prioritization of voice and video flows. We've been seeing a move towards this kind of thing for years. With things like the Aruba WiFi solutions. You may have 1000 access points or APs, but you don't manage 1000 APs, right? You manage a few mobility controllers, perhaps all overseen by a mobility conductor, all of which helps to automate management functions. And this trend continues with the ability to move those mobility conductors into the cloud, giving you evermore centralized services. And we can leverage this functionality for our wired LAN solutions as well. With solutions like Aruba, OS, CX and Aruba's central management solutions, Aruba has been working towards moving all their service functions into an edge cloud delivery model. By doing this, they can deliver services ranging from automated RF planning and optimization in production, upgrade service and energy saving services, contract tracing and more. Okay, that concludes the NMS discussion evolution with tighter integration, AI and ML and the edge cloud delivery model. I hope you come back because this next topic is a good one. It's all about security
