What i am trying to do is to collect statistical data on server side for monitoring purpose. So when a request is hit to server, web interfaces refreshes its topology tab and we can see some statistics above Average Latency | 95 %tile | Throughput. So one thing I wish to know is what is the significance of these numbers and when I click on them I get some statistics right hand side of screen, which I mentioned in my first post.
Can you explain them and suggest some more ways for monitoring client requests, if possible.
The detail panel includes two groups of metrics, titled Interval Latency Measurement and Cumulative Latency Measurement.
The interval group shows metrics recorded during the most recent collection interval. The length of this interval is set by the RepNode parameter statsInterval, and it can also be seen as the difference between the start and end timestamps in the interval group.
The cumulative group shows metrics recorded over a longer period, generally since the RepNode was started, although some events such as changing RepNode parameters can cause the start time to be reset.
Each group contains these fields:
<li> totalOps - Simply the number of operations performed during the interval
<li> throughput - Operations per second during the interval
<li> min, max, avg - These are, respectively, the smallest, largest, and average latencies observed during the interval, in ms
<li> percent95, percent99 - These are the average latencies of the 95th and 99th percentile of operations.
You can set limits for the interval values of throughput and avg by setting the Admin parameters throughputFloor and latencyCeiling. If these limits are exceeded during the measurement interval, then the measurement is marked as alertable. When an alertable measurement event is received at the Admin, if a browser is connected to the Admin Console web application, then the browser will pop up a dialog box with a warning that the constraint was violated.