System and method of applying power laws in optimizing network traffic
View Patent ↗A system and method of predicting network data traffic includes coupling a first group of clients to a current server that results in a current CPU utilization of the current server. A second group of clients are coupled to the current server. A load multiple is determined and the current CPU utilization is compared to a predicted CPU utilization. A server requirement is increased if the current CPU utilization is greater than or equal to the predicted CPU utilization.
1. A method of predicting network data traffic comprising:
coupling a first plurality of client nodes to a current server resulting in a current CPU utilization of the current server;
coupling a second plurality of client nodes to the current server;
determining a load multiple;
comparing the current CPU utilization of the current server to a predicted CPU utilization equal to an inverse of a product of the first plurality of client nodes and the load multiple to a scaling exponent of between about ⅓ and about ⅔ power; and
increasing a server requirement if the current CPU utilization of the current server is greater than or equal to the predicted CPU utilization.
2. The method of claim 1 , wherein the load multiple is equal to a sum of the first plurality of client nodes and the second plurality of client nodes divided by the first plurality of client nodes.
3. The method of claim 1 , wherein the scaling exponent is equal to about ⅔ for text file transfers.
4. The method of claim 1 , wherein the scaling exponent is equal to about ⅓ for all file transfers.
5. The method of claim 1 , wherein increasing the server requirement includes adding additional server CPU capacity.
6. The method of claim 1 , wherein coupling a first plurality of client nodes to the current server includes receiving a first plurality of requests from the first plurality of client nodes and wherein coupling a second plurality of client nodes to the current server includes receiving a second plurality of requests from the second plurality of client nodes.
7. The method of claim 1 , wherein increasing the server requirement includes outputting the predicted CPU utilization.
8. A method of predicting network data traffic comprising:
coupling a first plurality of client nodes to a current server;
adding a second plurality of client nodes to the current server;
determining a load multiple that is equal to a sum of the first plurality of client nodes and the second plurality of client nodes divided by the first plurality of client nodes;
comparing a current CPU utilization of the current server to a predicted CPU utilization equal to an inverse of a product of the first plurality of client nodes and the load multiple to a ⅓ power; and
increasing a server requirement if the current CPU utilization of the current server is greater than or equal to the predicted CPU utilizaton.
9. A system for managing network data traffic comprising:
a server system coupled to a computer network;
a first plurality of clients coupled to the network;
a second plurality of clients coupled to the network;
a load managing device coupled to the network, the load managing device including:
logic that determines a current CPU utilization of the server system;
logic that determines a load multiple;
logic that compares the current CPU utilization of the server system to a predicted CPU utilization equal to an inverse of a product of the first plurality of client nodes and the load multiple to a scaling exponent of between about ⅓ and about ⅔ power; and
logic that increases a server requirement if the current CPU utilization of the server system is greater than or equal to the predicted CPU utilization.
10. The system of claim 9 , wherein the load multiple is equal to a sum of the first plurality of clients and the second plurality of clients divided by the first plurality of clients.
11. The system of claim 9 , wherein the scaling exponent is equal to about ⅔ for text file transfers.
12. The method of claim 9 , wherein the scaling exponent is equal to about ⅓ for all file transfers.
13. The system of claim 9 , wherein logic that increases the server requirement includes logic that adds additional server CPU capacity.
14. The system of claim 13 , wherein logic that increases the server requirement includes logic that couples at least one additional server to the network.
15. The system of claim 9 , wherein the load managing device includes:
logic that receives a first plurality of requests from the first plurality of clients; and
logic that receives a second plurality of requests from the second plurality of client nodes.
16. The system of claim 9 , wherein the logic that increases the server requirement includes logic that outputs the predicted CPU utilization.