Introduction

WebMeter provides performance data on two key web metrics.  These are bandwidth and latency.  Bandwidth is important as it is the major factor in how long it takes to upload and download data.  Latency is important as it is the major factor in the responsiveness of your browsing experience (or SaaS functionality). 

The WebMeter performance report provides a dashboard showing a summary of these performance metrics, a history of the recorded performance over time and a detailed table of the data that forms the basis of the performance report.  To create a performance report follow these steps:

1. Follow the instructions in Creating Reports to get to the WebMeter report start page.
2. Optionally, fill out any parameters that you require according to the instructions here.
3. Click on the "Web Performance" button as highlighted in Figure 1, below.
   

Dashboard

The dashboard tab, illustrated in Figure 2, provides best, worst and average statistics for download performance, upload performance and the latency that you experienced within the scope of the report parameters.

Download performance is measured in Mbit/s.  Each gauge for this statistic (best, worst and average) has a red, amber and green color scheme to indicate the level of performance achieved.  In Figure 2 below, we can see that the best download performance achieved was 5.63 Mbit/s out of a potential maximum of 8 Mbit/s.  This maximum was specified during installation and can be configured as your ISP or network connections change.

You will note from Figure 2 that the worst download performance was 0 Mbit/s.  This can happen when WebMeter did not have appropriate data to calculate the statistic or the figure was so low that it does not register on the scale being used (i.e. Mbit/s).



Upload performance is measured in Mbit/s.  Each gauge for this statistic (best, worst and average) has a red, amber and green color scheme to indicate the level of performance achieved.   WebMeter's gauges need to cover both symmetrical (e.g. SDSL) and asymmetrical (e.g. ADSL) connectivity so does not alter the red, amber, green scales on upload to reflect the reduced capacity that is usual in domestic ADSL connections.  In Figure 2 above, we can see that the best upload performance achieved was 0.16 Mbit/s out of a potential maximum of 8 Mbit/s.

Latency is measured in milliseconds.  Each gauge for this statistic (best, worst and average) has a red, amber and green color scheme to indicate the level of performance achieved.  The scale used for this colour scheme is based on reasonable thresholds that hold for the vast majority of Internet connectivity.  In the example of Figure 2 a respectable 78 milliseconds latency was the best achieved with the worst being 13 seconds.

History

The history tab, illustrated in Figure 3, provides a graphical view of how performance varied over time for download, upload and the latency within the scope of the report parameters.  A separate graph is presented for each of these three metrics, only download speed is fully illustrated in Figure 3.

Each graph allows you to drill into detail by clicking one of the "zoom" options in the top left hand corner.  As you can see in Figure 3 this zoom resolution goes from 1 minute to 5 minutes and all the way to 6 months and 1 year.  This can be useful when attempting to track down a performance issue which occurred at a particular point in the past.

Detail

The detail tab, illustrated in Figure 4, provides a detailed tabular of each web transaction that has been used to build up the previous two performance reports.  The columns in this table are as follows:

• Date - the date at which the transaction took place.
• Site - the web site which was involved in the request / response transaction.
• Latency - the latency experienced before a response began to be received.
• Download - the download bandwidth (Mbit/s) if the response data was of a sufficient size to support an accurate calculation.
• Response - the size of the HTTP response.
  
• Upload - the upload bandwidth (Mbit/s) if the request data was of a sufficient size to support an accurate calculation.
• Request - the size of the HTTP request.
• Status - the HTTP status that was reported from the site for the request.
• Path - the path that was requested from the web site.
  

The table is paginated so that you can conveniently click though transactions a batch at a time.  Also, it is possible to order transactions by individual column values by clicking on any column.  For example, you can order transactions by request size by clicking on the "Request" column.

At the bottom left hand corner of the table there is a button which allows you to export all data in the table as a CSV file.  On windows platforms, this will allow you to open data in Microsoft Excel for further analysis and visualisation.