In a (default) Diagnostic chart, if both the pre- and post-tests for bandwidth fail, the chart hides the test data.
In DataView pages and charts, if either the pre- or post-tests for bandwidth fail, the chart hides the test data.
The different approaches occasionally cause the total test counts to be different between pages. This is a known issue and is being reviewed to be addressed in a future release.
The following is general information on Last Mile testing that provides insight into pre- and post-tests for bandwidth that are reflected in Last Mile charts.
Before (pre) and after (post) every Last Mile test, Gomez measures the bandwidth of the computer running the test to categorize the test results by available bandwidth.
Gomez tests the pre- and post-test bandwidth by accessing a random object that is hosted at a third party provider. The object download is timed for calculation purposes only.
Gomez uses the object download times to calculate the current bandwidth for that machine and classify the machine by averaging the results of the pre- and post-tests. The data collected for the test transaction is then placed in one of the following seven bandwidth buckets:
Legacy/Saturated G0 (<=56kbps)
Narrowband G1 (>56kbps, <=200kbps)
Narrowband G2 (>200kbps, <=512kbps)
Broadband G3 (>512kbps, <=768kbps)
Broadband G4 (>768kbps, <=1536kbps)
Broadband G5 (>1536kbps, <=3072kbps)
Broadband G6 (>3072kbps)
If a bandwidth test is problematic, the results of that machine are displayed in the test results. In other words, the bandwidth results for that machine are marked as zero and data from this machine is not displayed in the Gomez Portal.
Occasionally, when a download test is problematic, the data is hidden from reporting. As stated above, there is a difference between the logic in (default) Diagnostic Charting and DataView pages for hiding the data.