Historical performance data of various resources is collected in a fixed-size binary file. The older data is overwritten as new data arrives in a circular round robin array. The number of metrics, frequency of data insertion, number of objects, and the roll-up databases affect the size of binary file. The higher resolution data is compressed to a lower resolution data. For example, data is collected every 5-minutes for the last 24 hours for daily performance analysis. For monthly performance analysis, average of the 5-minutes data for every two hours is used. For the yearly performance log, average of the 5-minutes data for every 24 hours is used.