Results by Location

Results by location

When a test runs from more than one location, the Overview adds a By location table under the metric tiles: for each load-generator location the number of samples, the average and maximum response time, the error rate and the requests per second it produced. The location with the highest average is highlighted.

What the differences mean

The tiles above the table mix every location together. The table separates them, which is what you need to answer "is it slow for everyone, or slow from somewhere?"

• One location much slower, same error rate - network distance. A user in that region sees that latency; a CDN or a regional deployment is the fix, not more capacity.
• One location slower and failing - that location's traffic is being throttled or routed to a different, weaker backend. Compare the errors by type for that location's requests.
• All locations equally slow - the bottleneck is in the application or its database, not the path to it.
• Fewer samples from one location than its share of users - that generator was slower to ramp, or saturated; check the Engine Health tab.

Per-location charts

The Timeline tab's View per location button draws every metric as one line per location, so you can see whether a spike came from everywhere or from one region. Locations are configured on the test's geographical settings.