Routing improvement to reduce impact of out-of-resource condition
Disclosed are methods, controllers, and network devices that reduce the impact of an out of resource event for a network device. In one example, a controller can analyze routing information to determine which of a plurality of prefixes are dispensable based on several factors, such as the specificity of the prefix, the frequency of use of the prefix, the geographic area through which the prefix passes, and the extent to which upstream network devices rely on the prefix. Thereafter, the controller can remove the dispensable prefixes to resolve the out of network condition. In another example, the network device itself performs the analysis of which prefixes are dispensable and communicates that information to the controller. In either scenario, the controller can reduce micro-loops and sub-optimal forwarding during an out of resource condition for the network device by removing prefixes that have a lesser impact on the network as a whole.
1 . A method of controlling a network device, the method comprising:
receiving, by a BGP controller, a communication from the network device including routing information of next-hop prefixes for forwarding network traffic by the network device;
analyzing, by the BGP controller, the routing information and comparing the routing information to routing information of other network devices to determine at least one prefix of the prefixes that, if removed from a routing table of the network device, would result in a lower amount of efficiency reduction for routing the network traffic in the network, the analyzing including:
associating, with at least some of the prefixes, one or more values derived from evaluating the routing information; and
evaluating the values according to one or more criteria;
receiving an out of resource notification from the network device indicating an out of resource condition of the network device; and
modifying a local policy of the network device to alleviate the out-of-resource condition by first removing the at least one prefix based on the evaluating the values according to the one or more of criteria.
2 . The method of claim 1 ,
wherein the value is a prefix discard value assigned to at least some of the prefixes using a discard logic configured to rank the prefixes based on an effect of the prefixes on the efficiency reduction;
wherein the evaluating the values according to one or more criteria includes determining whether the prefix discard value exceeds a discard value threshold; and
wherein modifying the local policy includes removing the prefix if the prefix discard value exceeds the discard value threshold.
3 . The method of claim 2 , wherein the discard logic assigns the prefix discard value based on at least one member selected from a group consisting of a specificity of the prefix, a frequency of use of the prefix, an availability of alternative prefixes for upstream network devices, a geographic area through which the prefix passes, and a use of the prefix by upstream network devices.
4 . The method of claim 1 , wherein the analyzing the routing information is performed more frequently as the communication from the network device indicates the network device is approaching the out of resource condition.
5 . The method of claim 1 , further comprising analyzing the communication from the network device to identify an upstream network device that routes through the network device, and communicating with the upstream network device to notify the upstream network device that the prefix has been removed.
6 . The method of claim 1 , wherein the out of resource notification indicates the network device has reached or exceeded a threshold of resource utilization of the network device.
7 . The method of claim 1 , wherein the step of receiving the out of resource notification is performed after the step of analyzing.
8 . A controller for managing a network including a network device, the controller comprising:
a storage configured to store instructions; and
at least one processor configured to execute the instructions and cause the at least one processor to:
receive, by a BGP controller, a communication from the network device including routing information of next-hop prefixes for forwarding network traffic by the network device;
analyze, by the BGP controller, the routing information and comparing the routing information to routing information of other network devices to determine at least one prefix of the prefixes that, if removed from a routing table of the network device, would result in a lower amount of efficiency reduction for routing the network traffic in the network, the analyzing including the steps:
associating, with at least some of the prefixes, one or more values derived from evaluating the routing information; and
evaluating the values according to one or more criteria;
receive an out of resource notification from the network device indicating an out of resource condition of the network device; and
modify a local policy of the network device to alleviate the out-of-resource condition by first removing the at least one prefix based on the evaluating the values according to the one or more of criteria.
9 . The controller of claim 8 ,
wherein the value is a prefix discard value assigned to at least some of the prefixes using a discard logic configured to rank the prefixes based on an effect of the prefixes on the efficiency reduction;
wherein the evaluating the values according to one or more criteria includes determining whether the prefix discard value exceeds a discard value threshold; and
wherein modifying the local policy includes removing the prefix if the prefix discard value exceeds the discard value threshold.
10 . The controller of claim 9 , wherein the discard logic assigns the prefix discard value based on at least one member selected from a group consisting of a specificity of the prefix, a frequency of use of the prefix, an availability of alternative prefixes for upstream network devices, a geographic area through which the prefix passes, and a use of the prefix by upstream network devices.
11 . The controller of claim 8 , wherein the analyzing the routing information is performed more frequently as the communication from the network device indicates the network device is approaching the out of resource condition.
12 . The controller of claim 8 , further comprising analyzing the communication from the network device to identify an upstream network device that routes through the network device, and communicating with the upstream network device to notify the upstream network device that the prefix has been removed.
13 . The controller of claim 8 , wherein the out of resource notification indicates the network device has reached or exceeded a threshold of resource utilization of the network device.
14 . The controller of claim 8 , wherein the step of receiving the out of resource notification is performed after the step of analyzing.
15 . A non-transitory computer-readable storage medium for automated management of a subscription service provided by a subscription provider, the computer-readable storage medium including instructions that, when executed by at least one processor, cause the least one processor to:
receive, by a BGP controller, a communication from the network device including routing information of next-hop prefixes for forwarding network traffic by the network device;
analyze, by the BGP controller, the routing information and comparing the routing information to routing information of other network devices to determine at least one prefix of the prefixes that, if removed from a routing table of the network device, would result in a least amount of efficiency reduction for routing the network traffic in the network, the analyzing including:
associating, with at least some of the prefixes, one or more values derived from evaluating the routing information; and
evaluating the values according to one or more criteria;
receive an out of resource notification from the network device indicating an out of resource condition of the network device; and
modify a local policy of the network device to alleviate the out-of-resource condition by first removing the at least one prefix based on the evaluating the values according to the one or more of criteria.
16 . The computer-readable storage medium of claim 15 ,
wherein the value is a prefix discard value assigned to at least some of the prefixes using a discard logic configured to rank the prefixes based on an effect of the prefixes on the efficiency reduction;
wherein the evaluating the values according to one or more criteria includes determining whether the prefix discard value exceeds a discard value threshold; and
wherein modifying the local policy includes removing the prefix if the prefix discard value exceeds the discard value threshold.
17 . The computer-readable storage medium of claim 16 , wherein the discard logic assigns the prefix discard value based on at least one member selected from a group consisting of a specificity of the prefix, a frequency of use of the prefix, an availability of alternative prefixes for upstream network devices, a geographic area through which the prefix passes, and a use of the prefix by upstream network devices.
18 . The computer-readable storage medium of claim 15 , wherein the analyzing the routing information is performed more frequently as the communication from the network device indicates the network device is approaching the out of resource condition.
19 . The computer-readable storage medium of claim 15 , further comprising analyzing the communication from the network device to identify an upstream network device that routes through the network device, and communicating with the upstream network device to notify the upstream network device that the prefix has been removed.
20 . The computer-readable storage medium of claim 15 , wherein the out of resource notification indicates the network device has reached or exceeded a threshold of resource utilization of the network device.