Ask any storage administrator trying to deploy Business Continuity and Disaster Recovery (BC/DR) solutions between two data centers if it is simple  to get the networking set up for mission critical applications in a BC/DR scenario.  I am willing to wager you'll hear the same tales I've often heard, that the storage team and the network team have a difficult time setting up the necessary network configurations for the storage replication application's service level requirements between sites.  Equally important, there is also a complicated challenge with providing a duplicate infrastructure at the secondary site so that the application will be able to come up to restore the application to service  quickly (often referred to as the Recovery Time Objective, RTO).

Cisco's Application Centric Infrastructure enables a novel solution to this problem which one of our ACI partners, World Wide Technology, Inc. (WWT), recently deployed. It illustrates the power of the Application Policy Infrastructure Controller's (APIC) and the ACI fabric's open programmability. To appreciate this, let's take a closer look at the problem.

There are cloud-based solutions to synchronize virtual machines between data center sites and storage solutions to synchronize databases and replicate files systems between sites. IT operations teams deploying BC/DR solutions fully understand the need replicate the application networks between the sites that are supporting the application as well.

Imagine you have a primary cloud data center with a vital multi-tier application which needs to keep running 99.9% of the time. Also assume that your primary site is dynamically adjusting compute, network and storage resources frequently in response to varying demand. At the secondary data center, your network operations teams have to maintain matching layer 2 and layer 3 network connectivity and L4-L7 network services for the application with the changes made at the primary data center. Replicating policy automatically at the network layer between data center sites is something new that a policy-driven data center makes possible.

Since the APIC exposes a REST interface that allows customers to add functionality on their own, WWT wrote a custom APIC solution to replicate the network policy for a tenant or an application between two data center sites to solve this common problem.   Read Ravi Balakrishnan's blog on today's Cisco Data Center blog for more insight into how they did this.

In this age of cloud, big data, social and mobile applications, enterprises find an ever growing majority of their revenues are dependent on the non-stop operation of their applications, data, and networks in their data centers.  As this solution from WWT shows, ACI programmability has enabled an innovative multi-vendor solution to the problem of replicating the application -its virtual machines, virtual disks, and supporting application network between data center sites.

Links

Cisco Data Center Blog: Federated ACI Fabrics for Dual Data Center Deployments -A Disaster Recovery ACI use-case

ACI Disaster Recovery Demonstration

WWT White Paper: Federated Application Centric Infrastructure (ACI) Fabrics for Dual Data Center Deployments

Cisco Go/ACI web page