Distributed Cloud Computing
A Google Data center in Alaska is an attractive proposition to our community and would certainly enhance the success of regional initiatives, including the innovative applications we are presenting in our proposal. A local data center would benefit from inexpensive land, power and cooling.
At the same time, massive internet data are not as cost effective, and certainly not as agile, as service provisioning using existing resources. Use of distributed resources achieves a balance of resource utilization which effectively complements the symmetric network service provisioning envisioned by both Google and the Anchorage partners.
The critical aspect of Distributed Cloud Computing is the extension of the footprint of the cloud onto all participating resources in the network. With that capability in place a number of interesting service provisioning models could be experimented with, from a fully distributed Google Cloud to a Community cloud which incorporates elements of a peer to peer model with a more distributed control mechanism.
Latency and backhaul issues which characterize Alaska traffic engineering suggest that Anchorage would be a perfect location to demonstrate the benefits of a fully distributed cloud computing service model.
If however Google is interested in deploying a local data center to ensure the success of the Google Fiber for Communities program for our region we believe the low cost of land, power and cooling make the Anchorage area an attractive locale to build and operate Google’s critical infrastructure.
