95th Percentile and Moving Average calculation

95th Percentile and Moving Average calculation

95th Percentile

The 95th percentile is a number that is greater than 95% of the numbers in a given set. The reason this statistic is so useful in measuring data throughput is that it gives a very accurate picture of the maximum traffic generated on an interface. This is a standard measure used in interpreting performance data. This 95th percentile is the highest value left when the top 5% of a numerically sorted set of collected data is discarded. It is used as a measure of the peak value used when one discounts a fair amount for transitory spikes. This makes it markedly different from the average.

The 7/30 Reports is available for various attributes/resource of a particular monitor. Apart from displaying availability for the past 7 days or 30 days, the 7/30 reports also display the reports for each attribute/resource, which can be exported into PDF formats.

To calculate the 95th percentile value,

  • Note the number of values in the sample (K).
  • Find N by calculating 95% of K. (N = K x 0.95)
  • Sort the samples in ascending order. Now, the Nth value in the sorted list will be the 95th percentile value.
Let us consider an attribute history report of CPU Utilization metric attached below. In this report, we have a total of 19 samples of CPU Utilization values. 


  • The total number of entries (K) in the report is 19. So, K = 19.
  • Since the 95th percentile value (N) is being calculated, multiply the number of entries (K) with 0.95.
    Therefore, N = 19 x 0.95 = 18.05
  • Now sort the values of all the entries in ascending order. The 95th Percentile value would be the 18th value (Nth value) in the sorted list, which is 33.167%.

Moving average

  1. Moving average is a statistic that captures the average change in a data series over time.
  2. It is an unweighted mean of previous n data points over a stipulated time period, which are then used to spot performance trends in a graph.
  3. This graph is helpful in plotting trends easily by smoothing out the minor fluctuations in the performance data and makes forecasting easier. For example, a 10-day simple moving average of attribute value like CPU utilization is the mean of the previous 10 days' CPU Utilization value.
  4. You can add moving average graph in the 7, 30 graphs of the various monitors.
  5. By default, 1 day (24 hours) window is considered for calculating Moving Average values, which can be modified by specifying the required value for the option Show Moving Average Values based on last ___ Days located under Settings -> Report Settings.

                    New to ADSelfService Plus?

                      • Related Articles

                      • Calculation of various attributes in Linux server

                        Physical memory utilization Command: free -m Sample Output: total used free shared buff/cache available Mem: 16338247680 6590242816 1372995584 662192128 8375009280 8734982144 Swap: 100 10 0 Calculation: Physical memory % = 100 -(((MemFree + Buffers + ...
                      • Calculation of various attributes in AIX server

                        Page Space Command: lsps -a Sample Output: Page Space      Physical Volume   Volume Group       Size       %Used       Active     Auto        Type       Chksum ...
                      • Memory Usage Calculation (Linux- Telnet/SSH)

                        Total Memory Utilization Calculation: Linux Monitoring: Telnet/SSH For memory usage calculation using 'free -m'  command. Sample command output: root@root:~$  free -m                      total        used        free      shared  buff/cache   ...
                      • Is Average Queue Length metric available in AIX server monitoring?

                        No, Average Queue Length is not supported in AIX server monitoring.
                      • Calculation of various attributes in MACOS server

                        Memory Utilization Command: export UNIX95;top -l 1 -n 2 Sample Output: Processes: 154 total, 2 running, 152 sleeping, 600 threads  2020/03/04 14:49:04 Load Avg: 1.46, 1.60, 1.62  CPU usage: 5.40% user, 8.10% sys, 86.48% idle  SharedLibs: 202M ...