Full Scale Simulated Network
The connectivity side of the house just completed our first full scale simulated network.
It consists of a main datacenter and a disaster recovery datacenter, each with two redundant network routers as well as two redundant network switches. There are several additional routers per datacenter to handle ancillary network services including a DMVPN buildout, and “fake internet”.
There are two branch offices each with two network switches to connect “fake PCs” and two routers to connect each branch office back to the datacenter by way of simulated private WAN circuits.
There are four remote offices each with a router and switch and connect to each datacenter by way of Warroom, our DMVPN cloud and uses “fake internet” as transport.
The private WAN circuits are supplied by our own network transport core that emulates circuit delivery the same way we currently provide network services to our clients. This NXP core is brought to life via 12 additional virtual routers and switches.
There are currently 19 virtual servers running on both Windows and Linux providing a number simulated system resources such as file sharing, SQL, Web, DNS, API and more. These servers are placed at the datacenters, branch offices as well as one remote offices to demonstrate the flexibility of our true life platform designs. We even have a server template created so adding a new server takes just a few seconds.
Each datacenter has two configured Cisco ASA security devices to simulate firewall and VPN needs. These devices too are connected to our “fake internet”
The fake internet currently consists of six fake providers that peer with eachother at a simulated internet exchange that was designed to mimic DACS, SPG’s IX current live in multiple regions across the country. This setup may be “overkill” but it will help serve as a proof of concept platform to help understand the need for peering as well as a training platform for DACS clients. Although we are currently at six fake providers, that number can be expanded to a hundred or more to truly provide global scale.
One real win for this environment is that it is monitored and maintained by several of our network application program products including Libre Network Management, Netbox Documentation Management, Oxidized / Rancid configuration management and alerting and a custom script platform developed completely in house that we call Vista. These demo systems are currently live, are internet accessible and can help clients see the value that these systems can bring at a fraction of the cost of main stream commercial competitive offering.
Another huge win for this network simulation environment is that much like software development it allows us to test new equipment and cutting edge networking concepts without effecting production. It can also assist in upgrade planning. They can also help validate production environment redundancy. With a click of a mouse, you can turn off a device and see how the network performs and mitigate around the problem area. These things are helpful to us, but as we found with our scripting engines it could also be a value proposition to our clients.
56 total network devices, a NXP core, a warroom core, a fake internet connected to a fake internet exchange, 19 simulated servers, connected and monitored to live web monitoring, reporting, and scripting tools....all on one piece of server hardware that we are barely consuming a quarter of.
It doesn’t take million dollar budgets to make dreams become a reality. It just takes a dream, and the desire to reimagine the impossible. It’s kinda fun to do the impossible!
Have a virtual look!