Dynamic load balancing using quality/loading bands

Abstract

Methods and apparatus for load balancing data traffic are disclosed. An example method includes determining a respective quality metric for each of a plurality of members of an aggregation group of the network device, each respective quality metric representing respective data traffic loading for each member of the plurality of aggregation group members. The example method also includes grouping the plurality of aggregation members into a plurality of loading/quality bands based on their respective quality metric. The example method further includes selecting members of the aggregation group for transmitting packets from a loading/quality band corresponding with members of the aggregation group having lower data traffic loading relative to other members of the aggregation group.

What is claimed is: 
     
         1 . A method comprising:
 determining, by a network device, a respective quality metric for each of a plurality of members of an aggregation group of the network device, each respective quality metric representing respective data traffic loading for each member of the plurality of aggregation group members;   grouping the plurality of aggregation members into a plurality of loading/quality bands based on their respective quality metric; and   selecting members of the aggregation group for transmitting packets from a loading/quality band corresponding with members of the aggregation group having lower data traffic loading relative to other members of the aggregation group.   
     
     
         2 . The method of  claim 1 , wherein each respective quality metric is based on at least one of port loading corresponding with each of the plurality of aggregate members and an amount of queued data corresponding with each of the plurality of aggregate members. 
     
     
         3 . The method  claim 1 , wherein each respective quality metric is determined by combining a first quality metric corresponding with a respective port loading and a second quality metric corresponding with a respective amount of queued data. 
     
     
         4 . The method of  claim 1 , wherein selecting members of the aggregation group includes randomly selecting members from the plurality of aggregation group members. 
     
     
         5 . The method of  claim 1 , wherein selecting members of the aggregation group includes deterministically selecting members from the plurality of aggregation group members. 
     
     
         6 . The method of  claim 1 , wherein selecting members of the aggregation group includes selecting members from the plurality of aggregation group members using round-robin selection. 
     
     
         7 . The method of  claim 1 , wherein the loading/quality bands are contiguous and non-overlapping. 
     
     
         8 . The method of  claim 1 , wherein each of the plurality of members comprises a plurality of physical links. 
     
     
         9 . The method of  claim 1 , wherein each of the plurality of members comprises a plurality of virtual links. 
     
     
         10 . The method of  claim 1 , wherein determining the respective quality metric for each of the plurality of members of the aggregation group includes determining a respective quantized quality metric based on at least one threshold value. 
     
     
         11 . A network device comprising:
 an imbalance detection module configured to:
 determine a respective quality metric for each of a plurality of members of an aggregation group of the network device, each respective quality metric representing respective data traffic loading for a corresponding member of the aggregation group; and 
   an aggregate member selection module coupled with the imbalance detection module, the aggregate member selection module being configured to:
 group the plurality of aggregation members into a plurality of loading/quality bands based on their respective quality metric; and 
 select members of the aggregation group for transmitting packets from a loading/quality band corresponding with members of the aggregation group having lower data traffic loading relative to other members of the aggregation group. 
   
     
     
         12 . The network device of  claim 11 , wherein the imbalance detection module is configured to determine the respective quality metric for each member of the plurality of aggregation group members by determining a respective quantized quality metric based on at least one threshold value. 
     
     
         13 . The network device of  claim 11 , wherein the imbalance detection module is configured to determine the respective quality metric for a given member of the plurality of aggregation group members by combining a first quality metric corresponding with a respective port loading for the given member and a second quality metric corresponding with a respective amount of queued data for the given member. 
     
     
         14 . The network device of  claim 11 , wherein the loading/quality bands are contiguous and non-overlapping. 
     
     
         15 . The network device of  claim 11 , wherein the aggregate member selection module is configured to select members of the aggregation group by randomly selecting members from the plurality of aggregation group members. 
     
     
         16 . The network device of  claim 11 , wherein the aggregate member selection module is configured to select members of the aggregation group by deterministically selecting members from the plurality of aggregation group members. 
     
     
         17 . The network device of  claim 11 , wherein the aggregate member selection module is configured to select members of the aggregation group by selecting members from the plurality of aggregation group members using round-robin selection. 
     
     
         18 . A method for determining a quality metric for a given aggregate member of a plurality of aggregation members in a network device, the method comprising:
 periodically determining, by the network device, an amount of average data traffic loading for the given aggregate member;   quantizing the amount of average data traffic loading to a binary loading metric;   periodically determining an amount of average queued data for the given aggregate member;   quantizing the amount of average queued data for the given aggregate member to a binary queue size metric;   combining the binary loading metric and the binary queue size metric to produce a combined quality metric.   
     
     
         19 . The method of  claim 18 , further comprising mapping the combined quality metric to a respective quality measure using a randomly selected quality mapping function. 
     
     
         20 . The method of  claim 18 , further comprising mapping the combined quality metric to a respective quality measure using a lookup table to map the combined quality metric to the respective quality measure.