The Trick To High Availability Services (Cont'd)

A better solution is an IP management system that understands the hierarchy of an end-to-end service construction including intimate knowledge of the component technologies that comprise it. This “service awareness” allows the management system to react to individual nodal and composite service events and present correlated, relational information to the operator, allowing faster service provisioning, verification, and restoration. Specifically, service aware templates are created which understand not just the individual services and elements that might be affected by the provisioning of a new service, but also the connection points between technologies that comprise the composite service as it traverses the network. For example, when a service is turned up at an edge port, the service aware manager knows how to construct and enable it, including paths, priorities, and other attributes. This creates an "auto-instantiating infrastructure" where the network builds itself through an over-arching understanding of the composite-service hierarchy. This simplifies and accelerates manual or OSS-driven provisioning and allows the integration of OAM policies with provisioning for automated turn up testing.

The challenge of increasing IP service availability is in maintaining a stable, accurate routing environment. Routers make forwarding decisions for customer traffic based on network topology information each node continually updates and maintains. This is a very dynamic and complex function requiring constant exchange of information amongst routers. The challenge is maintaining this information when a router switches between primary and backup control processors. The established norm for Internet routers, however, is for routing protocols to stop and restart.

Euphemistically called Graceful Restart, the affected router enters a recovery mode during which its routing protocols reconverge their understanding of the network topology. During this process, forwarding decisions for customer traffic is based on information that may no longer be valid, leading to service outages. Convergence of always-on services over IP demands more, and for this challenge, there is Non-Stop Routing. As its name implies, it is a technology whereby routing protocols remain fully operational during control plane switchovers, greatly reducing the amount of time topology changes are not able to be updated and reflected in forwarding decisions.

Another consideration is the impact routing protocol failures have on network reliability. The upside of dynamic routing protocols is that their interaction between network elements reduces operational cost. The downside, however, is that a problem in one network element can spread to a much larger population of elements, and this has been observed in both packet routing and TDM voice-switching networks. Non-Stop Routing greatly diminishes this possibility by isolating the corrective actions following a control plane processing failure to within the system where the failure occurred. This is particularly important in networks that have a mix of legacy and next generation routers where additional load on neighbor routers could have unintended consequences.

Massive, rapid deployment of value-added, “always-on” consumer and enterprise services has become a "do or die" imperative for operators. New converged IP service deliver architectures built with service aware management and Non-Stop Routing will unshackle operators from IP’s service complexity and "best effort" service heritage. And that’s something service providers can bank on to deliver high speed rollouts of always-on services.

« Previous   |   1   |   2   |   Subscribe

Send Comment

© 2005, All information contained herein is the sole property of Pipeline Publishing, LLC. Pipeline Publishing LLC reserves all rights and privileges regarding the use of this information. Any unauthorized use, such as copying, modifying, or reprinting, will be prosecuted under the fullest extent under the governing law.