plebanks.blogg.se - Using arista switch as a tor switch with scvmm

Each ACIsite has a single APIC cluster managing spine leaf fabric.

Enhanced scalability for a large data centre footprintĪCI Multi-site is two or more fabrics (having its own APIC cluster) that are managed as a unit using ACI multi-site orchestrator.

Ease of administration – All leaf and Spine switches deployed across Pod come under one Single APIC cluster which means they are considered a single administrative domain.

Control plane between Pod leverages MP-BGP EVPN so endpoint information is propagated in one Pod to an endpoint to another Pod in a seamless manner. All inter-Pod traffic is encapsulated with VXLAN.

Connectivity and control – all Pods within topology are interconnected using an IP routed inter-Pod network (IPN) which is not managed by APIC but user can configure it separately.

Administrative overhead is minimal while extending data centre network

Multi-Pod offers resiliency at network level across Pods and rest of the functionality remains with Single ACI fabric.

It is an evaluation of what was earlier called as ‘stretched fabric’. ACI Multi-pod fabric comprises two to twelve ACI Pods which are connected via an inter Pod network and managed under a single APIC cluster. The ACI multi-Pod architecture is an extension of pre-existing ACI fabric without the need to set up new fabric from the start. A ‘Site’ is a single fabric in the ACI world. If you have multiple fabrics, each with its own APIC cluster then independent ACI clusters are referred to as ‘Sites’. A ‘Pod’ is a set of interconnected ACI leaf and spine switches that are under the control of a specific APIC cluster.ĪCI fabric could have multiple Pods and all these Pods are part of the same fabric and are under the control of the same APIC cluster.

A ‘Fabric’ is a spine leaf topology of Nexus 9000 series switches with a single cluster of application policy infrastructure controllers (APIC) it is a single point of management for ACI fabric. Before we deep dive into these two distinct architectures we need to understand two terminologies commonly used or associated with them.