Abstract: One embodiment of the invention comprises a method for constructing a graph pertaining to a specified metric. Measured values of the specified metric are selectively stored, wherein each stored value comprises a data point. The method further includes routing successive data points to a location associated with the graph, wherein a first number of data points are included in a specified graph time period. The method further includes using a value that represents a particular data level of detail to divide the specified graph time period into multiple time intervals, wherein each time interval includes a second number of data points. The method further includes selectively processing the second number of data points of a time interval, in accordance with prespecified criteria, to determine a most representative value for that time interval. The most representative values of respective time intervals are then used to construct the graph.

Abstract: One embodiment of the invention comprises a method for constructing a graph pertaining to a specified metric. Measured values of the specified metric are selectively stored, wherein each stored value comprises a data point. The method further includes routing successive data points to a location associated with the graph, wherein a first number of data points are included in a specified graph time period. The method further includes using a value that represents a particular data level of detail to divide the specified graph time period into multiple time intervals, wherein each time interval includes a second number of data points. The method further includes selectively processing the second number of data points of a time interval, in accordance with prespecified criteria, to determine a most representative value for that time interval. The most representative values of respective time intervals are then used to construct the graph.