Connection distribution within a connectional parallelism architecture

Abstract

Example embodiments provide various techniques for distributing connections within a connectional parallelism architecture. In one embodiment, a method is provided where resource utilizations of connection groups are measured. Here, each connection group is assigned to one of multiple processors. A probability distribution is accessed that maps probabilities assigned to relative resource utilizations. A relative resource utilization of one of the connection groups is determined based on a resource utilization of the one connection group relative to other resource utilizations of other connection groups. A probability from the probability distribution is identified based on the determined relative resource utilization, and based on the identified probability, a connection is assigned to this connection group for execution by one of the processors assigned to this connection group.

What is claimed is: 
     
       1. A method of distributing connections within a connectional parallelism architecture, the method comprising:
 accessing a record of previous connection group assignments for multiple connection groups; 
 selecting one of the connection groups for a connection based on the record; 
 determining a current relative resource utilization of the selected connection group as compared to other resource utilizations of other connection groups; 
 identifying a probability from a probability distribution based on the determined current relative resource utilization; and 
 based on the identified probability, assigning the connection to the selected connection group for execution by a processor assigned to the selected connection group; and 
 
       wherein the probability from the probability distribution decreases as the determined current relative resource utilization increases. 
     
     
       2. The method of  claim 1 , wherein selecting one of the connection groups for the connection includes a group preference for the connection, the group preference defining a connection group preferred by the connection over other connection groups. 
     
     
       3. The method of  claim 2 , wherein the group preference is predefined. 
     
     
       4. The method of  claim 2 , wherein the group preference is defined based on a type of Transmission Control Protocol (TCP) connection. 
     
     
       5. The method of  claim 1 , wherein each connection group is identified by a group identifier, and wherein the assignment of the connection comprises assigning the connection to a group identifier that identifies the selected connection group. 
     
     
       6. The method of  claim 1 , wherein the probability distribution is predefined, and wherein the probability distribution corresponds to a step function-based distribution in which probabilities decrease as relative resource utilizations increase. 
     
     
       7. The method of  claim 1 , wherein the resource utilizations of the multiple connection groups are determined by measuring a number of connections to each of the multiple connection groups. 
     
     
       8. The method of  claim 7 , wherein the current relative resource utilization is determined based on a ratio of a number of connections owned by the selected connection group to a total number of connections. 
     
     
       9. The method of  claim 1 , wherein the resource utilizations of the multiple connection groups are determined by measuring data rates serviced by the multiple connection groups. 
     
     
       10. The method of  claim 9 , wherein the current relative resource utilization is determined based on a ratio of a data rate serviced by the selected connection group to a total data rate serviced by the connection groups. 
     
     
       11. A system comprising:
 a plurality of processors including at least a first processor and a second processor; and 
 a memory in communication with the first and second processors, the memory being configured to store a connection distribution module having instructions that when executed, cause operations to be performed, the operations comprising:
 accessing a record of previous connection group assignments for multiple connection groups; 
 selecting one of the connection groups for a connection based on the record; 
 determining a current relative resource utilization of the selected connection group as compared to other resource utilizations of other connection groups; 
 identifying a probability from a probability distribution based on the determined current relative resource utilization; and 
 based on the identified probability, assigning the connection to the selected connection group for execution by a processor assigned to the selected connection group; and 
 wherein the probability from the probability distribution decreases as the determined current relative resource utilization increases. 
 
 
     
     
       12. The system of  claim 11 , wherein selecting one of the connection groups for the connection includes a group preference for the connection, the group preference defining a connection group preferred by the connection over other connection groups. 
     
     
       13. The system of  claim 12 , wherein the group preference is predefined. 
     
     
       14. The system of  claim 12 , wherein the group preference is defined based on a type of Transmission Control Protocol (TCP) connection. 
     
     
       15. The system of  claim 11 , wherein each connection group is identified by a group identifier, and wherein the assignment of the connection comprises assigning the connection to a group identifier that identifies the selected connection group. 
     
     
       16. The system of  claim 11 , wherein the probability distribution is predefined, and wherein the probability distribution corresponds to a step function-based distribution in which probabilities decrease as relative resource utilizations increase. 
     
     
       17. The system of  claim 11 , wherein the resource utilizations of the multiple connection groups are determined by measuring a number of connections to each of the multiple connection groups. 
     
     
       18. A non-transitory, machine-readable medium storing instructions that, when performed by a machine, cause the machine to perform operations comprising:
 accessing a record of previous connection group assignments for multiple connection groups; 
 selecting one of the connection groups for a connection based on the record; 
 determining a current relative resource utilization of the selected connection group as compared to other resource utilizations of other connection groups; 
 identifying a probability from a probability distribution based on the determined current relative resource utilization; and 
 based on the identified probability, assigning the connection to the selected connection group for execution by a processor assigned to the selected connection group; and 
 wherein the probability from the probability distribution decreases as the determined current relative resource utilization increases.