[Note: Register today for our upcoming live ACI webcast: “Is Your Data Center Ready for the Application Economy”, January 13, 2015, 9 AM PT, Noon ET, featuring ACI customers and several key ACI technology partners.]
At the most recent Gartner Data Center Conference in Las Vegas, after some insightful discussions with customers and analysts, we came up with a great demo idea and proof point that highlights a key feature in our Application Centric Infrastructure (ACI) platform. This particular demo centers on the unique visibility of the ACI Fabric to faults in the underlying physical network.
Joe Onisick, Principal Engineer in the ACI team at Cisco, compares this ability in ACI to SDN technologies that employ only virtual overlay networks in the following video. With overlay networks, such as a VXLAN tunnel, the resulting virtual network (and all the management and analytics tools) has a much harder time isolating faults within the physical infrastructure. The overlay is designed to “tunnel” through the physical network, simplifying and obscuring the physical topology and issues with any specific network node. Before going much further, I’ll let Joe provide the details in this quick, 3 minute video:
https://www.youtube.com/watch?v=yZu-JW-DEQ8
In this demo scenario, Joe eliminates one of the network connections in the ACI fabric, affecting traffic distribution to applications running off that network leaf, a typical Day-2 operational challenge that resonates with IT teams. Because the Nexus 9000 switches in the ACI fabric are monitored by the ACI SDN controller, APIC, we can get immediate notification of the resulting network fault, and begin remediation steps through the APIC GUI to quickly isolate the issue to a particular node or nodes.
Health scores, a key attribute of a particular application network in the ACI model, can immediately signal the scope and severity of the problem, and provide guidance to the administrator about the remediation process. In an overlay environment, much of this information would be lost to the SDN monitoring tools.
Want more information? Register for our upcoming webcast featuring enthusiastic ACI customers. Or you can learn more about the design of the ACI fabric, the APIC controller and how its designed for zero-touch provisioning here. Or check out all of our Cisco ACI white papers here.
Cool, Nice demo …
Had a couple of question:-
1. How about visibility under hyper-visors eg : Veth going down, is this detected or reflected under APIC console ?
2. How about whole switch/leaf going down and how is that monitored and restored?
Regs Ravi
Ravi,
ACI has visibility into virtual network failures from any hypervisor platform. This would include a Veth going down. Additionally ACI has full visibility into the physical network fabric. ACI can immediately show not only that a switch or link has failed, but which switch/link and where packets are rerouting.
Joe Onisick – Principal Engineer – Cisco ACI/Nexus 9000
Nice Blog post, but just a small question:
How did you get the VMware Hands On Labs (http://hol.vmware.com) to have a link failure?