Reports#
What is kept after a run, and what you can send to somebody who was not watching.
Report generation#
Report → Report Generation. Give it a results file — from a run you just did, from history, or uploaded — and it produces the report: summary figures, response times over the run, throughput, errors, and a breakdown per request label.
Reports can be rendered as PDF, and a run can be set to email its report when it finishes; see Running a test.
Comparison reports#
Report → Comparison Reports puts two or more runs side by side. This is the one that answers "is it better than last time", which a single report cannot.
JTL Analyzer#
Report → JTL Analyzer reads a raw JMeter results file (.jtl) and produces the same figures the live panel showed, from the same source. Use it for a file produced outside JMXPress, or to re-read one later.
The figures are read from the results file, so they agree with what was on screen during the run rather than being a second calculation.
JTL Editor#
Admin → JTL Editor adjusts a results file — for example removing a warm-up period that should not count toward the figures.
Careful An adjusted file produces different numbers from the raw one. Adjustments are recorded so a report can say it was made from an edited file, but it is worth being explicit when you share one.
Prod Analyzer#
Utilities → Prod Analyzer reads production access logs rather than a test's results, to work out what the real traffic shape is — which endpoints, in what proportion, at what rate. That is the input to a load profile that resembles reality.
Little's Law#
Utilities → Little's Law relates concurrency, throughput and response time. Use it to sanity-check a load profile before running it: if the numbers you have chosen cannot produce the throughput you want, that is better known beforehand.
Memory, heap and thread analysis#
Covered in Heap dumps, thread dumps and logs.
APM metrics for a run#
A monitored run keeps the figures collected from your provider for its own window. See APM integrations.
What is kept, and for how long#
| Thing | Where it lives |
|---|---|
| Run history and figures | With the project the run belongs to |
| Results files | The library, under the run that produced them |
| APM metrics | With the run, for apm_days (default 90) |
| Diagnostics | For diagnostics_days (default 14) |
| Unsaved recordings | Swept after keep_recordings_days |
Note A run started from an upload rather than from a project has no project to be filed under. Running from a project is what makes a run findable afterwards.