Broadhop: Service velocity and scalability for policy

The announcement this week by Broadhop of the deployment of its policy server on Cisco’s Unified Computing System (UCS) enables the deployment of a “virtualized” policy layer by service providers, laying the ground work for potentially much more robust policy deployments.

For the most part, policy thus far has not really been done on a large scale; it’s been deployed mainly in more of a niche deployment manner. With 4G and next generation IP networks, however, operators will need larger-scale deployments, so policy will have to become massively scalable.

That has been an issue in the past, as so much redundancy is needed (especially at peak congestion times) for policy recognition and enforcement. Now data centers are becoming much more virtualized in general, service provides are looking to see virtualization moving more broadly into network and policy servers.

Because virtualization means that resources and information can be shared across networks and geographies, policy doesn’t have to be “server bound” any longer, so redundancy and load sharing can be automated. “You don’t have to have a single set of policies based on servers associated with a GGSN and put in multiple physical places around the network to take care of individual deployments,” said Dan Geiger, Broadhop senior director of marketing.

What this means is that subscribers and their profiles can be disseminated and shared at much faster speed so that all services associated with a certain customer can reside on multiple servers in a network, as opposed to having all information about one subscriber locked into one server on a GGSN. Instead, it can be spread across multiple resources so that profile information can be rapidly pulled down and services rapidly delivered or activated.

“We see a ten times increase in service delivery,” contended Geiger when talking about service velocity as a benefit of the virtualized solution.

Such an approach also provides more opportunity for rapid scalability and personalization as well. “Servers around the country can be used according to where the power is needed in an automated fashion,” said Geiger.