In part 1 we talked about how Cisco ACI simplifies diagnosis and enables DevOps Model compared to competing network virtualization solutions.
In this blog we’ll explore two additional critical scenarios that affect network IT teams. We’ll look at it from ACI perspective and compare that to other network virtualization solutions.
1) ACI Provides Proactive SLA Assurance
IT organizations are pressed to track and provide SLA especially for productivity and critical business applications. The challenge is typically around how to actively monitor SLA for specific app and how fast to go about improving SLA in case it drops below certain level?
In this scenario, IT wants the ability to monitor health score for a customer facing app and to be able to correlate performance degradation across the network to locate the problem. With ACI, you get real-time healthscore dashboard as well as built-in intelligence in the fabric to load balance application traffic as congestion occurs to meet SLA requirements. Also, the integration of L4-7 devices with APIC allows fault correlation across loadbalancers and firewalls to locate the problem.
In comparison to competing network virtualization solutions, they are unable to natively demonstrate real-time view of traffic/ application pattern, or automatically load balance traffic in case of congestion or link failure as well as limited integration with underlay to locate cause of problem.
See Joe Onisick explaining this here
2) ACI Designed For High Application Performance
In the new era of applications that require the highest performance from networks, we’ve seen leaf spine architectures becoming more common to optimize traffic patterns and minimize number of hops that packets traverse the network. So it is important the architecture is able to support both physical and virtual applications without complicated configurations or additional Layer 3 gateways that may impact performance.
In this scenario, users are experiencing application performance problem related to data residing on bare-metal DB server. Let’s look at how ACI architecture will handle this? First, the ACI fabric is automatically discovered and configured thus minimizing amount of time required to bring up switches and configure them. Second, every leaf switch which can be top-of-rack is a Layer 3 gateway for both physical and virtual workloads. Third, all are managed from a central location, APIC. Lastly, every packet is two hops away maximum from its destination within the data center and delivery of small packets won’t suffer in the presence of large packets.
This is unlike what customers may find in competing network virtualization solutions that require additional Layer 3 gateways. The result is likely to have impact on application performance, increased latency, and additional cost.
See Joe Onisick detailing this here