| 摘要 |
Systems and methods consistent with the present invention enable routing table updates are performed by optimally utilizing the resources of a node without exceeding the resources of the node. Using feedback on the amount of resources available to the nodes, such as in terms of available memory, the node may make new connections before breaking old one where those updates will not exceed available resources. This is referred to as make-before-break. When not enough resources are available, the node will break old connections before making new ones. This is referred to as break-before-make. Unlike the strict make-before-break and break-before-make models, this "loose" make-before-break method considers the amount of available resources in view of the resources required to perform the routing table updates without a node failure. Routes may also be tagged to prioritize the addition of more important routes and the deletion of less significant routes. Methods and systems consistent with the present invention, therefore, provide a routing table update method with which routing table updates are achieved without crashing and at the same time minimizing black hole intervals.
|
