Skip to main content
APIContext
Talk to salesStart free
Start HereGuidesConceptsCLIReferenceChangelog

Monitor Stats

Key Concepts​

APImetrics provides a variety of views of the statistics of each test that allows you to understand at a glance the behavior of each call.

As you deploy API test calls, APImetrics starts logging results. You can get to the stats for a specific test from the Quality Analytics view, a report, the home page, the list of calls or the call edit page.

The statistics for an individual test can be viewed either by all cloud locations by day, month or year, and by individual cloud location.

The views available are:

  • global latency analysis showing calls broken down by call portion and percentile
  • pass-fail rates - clicking on a day will take you to the filtered results for that day
  • latency over time
  • latency heatmap
  • latency histogram showing distribution of results
  • returned content size - useful for spotting content errors

Clicking on any view will take you to a larger version of it. You can also save latency graphs as reports.

Monitor Statistics overview showing Pass rate, Failure rate, Fastest and Slowest period, Latest Quality score, Latency Map, and Latency chart

Latency Map​

The analysis map gives you a geographical breakdown of total latency (or any specific latency component) by the locations of the data center (cloud location) the calls are being made from. You can zoom in on the map to see what is happening in a particular area more clearly.

1591

Guide to image:

  1. The larger a circle, the longer (slower) the call. Smaller circles are shorter (faster). Each circle is color-coded to the cloud it represents
  2. Metric - pick the monitor latency component you are interested in seeing
  3. Percentile Picker
  4. Domain Picker
  5. IP Address call made to
  6. Switch to logarithmic circles to make it easier to compare small and large results
  7. Cloud data center (location) picker
  8. Region picker
  9. IP Address - display results by IP address called - use in conjunction with 5)

Pass/Fail​

Pass/Fail block calendar showing green days (all passing) and a red cell on May 17 (some failures)

The Pass/Fail views gives you a quick overview of how the call has been behaving over the last few weeks. A green box indicates that there were no failures during that day. A red and green box indicates there were some failures during that day as well as passes. A red, yellow and green box indicates that there were some failures and warnings during that day as well as passes.

Latency Graph​

Latency graph showing average latency over the last month, fluctuating between 200–280 ms

The latency graph shows how the average latency has varied over the last few weeks.

Latency Heatmap​

Latency heatmap showing latency by day of week, color-coded from green (fast) to red (slow)

Latency Histogram​

Latency histogram showing the distribution of response times, with the majority of calls completing in the 200–300 ms bucket

Response Size​

642