Health
Testspace applies the same rules to every published Result: its Health, whether it is partial, the check it posts for CI, and what happens to metrics that stop reporting. This page explains those rules; Results and Metrics show where they appear. How a failing test is tracked from Result to Result is on Failures.
In brief:
- Health asks one question: did the checks you set pass? All passed: ✓ Healthy. Any failed: ✕ Unhealthy.
- "Partial" is information, not a verdict. It says the run was smaller than usual, or a metric that normally reports didn't. The verdict, statistics and CI check ignore it.
- Invalid is only for a run that can't be trusted: nothing ran, or more than 5,000 failures.
- CI build status is shown, never judged.
- A metric that stops reporting shows not reported for 5 Results, then is hidden until it reports again. Metrics Testspace created are deleted after 15 Results and 30 days without data.
Health
A Result's Health asks one question: did the checks you set pass? The checks are the test pass rate (the Space's Min pass threshold, 100% by default) and each enabled metric badge that has a threshold. The code coverage and static analysis metrics Testspace creates have none: they check nothing until someone sets a threshold or check, which their reports offer to do.
| Shown | When |
|---|---|
| ✓ Healthy | every check passed (or the Space has no checks set) |
| ✕ Unhealthy | a check failed |
| ! Invalid | nothing ran (no test cases and no metric values), or more than 5,000 cases failed, were blocked or exempt |
A Result with metric values but no test cases is judged on its metrics alone, and its pass rate shows a gray "—". If none of those metrics has a threshold, nothing is checked: the Result is Healthy, and its Health reads "No checks set". Invalid Results are left out of the Insights statistics.
Health drops and recovery. A Space's Health drops at an Unhealthy Result, and recovers at the next Healthy Result that ran every test still failing. A Healthy Result that skipped the failing tests (a subset that left out the failing Suite, say) doesn't end the drop. Insights reports the time to recover as Health Recovery Time.
Partial
Partial is a chip shown beside the verdict. It is information, not a judgement, and it says one or both of:
- The run was a subset: it had clearly fewer Suites (and cases) than the Space's recent Results; the Result shows how many ran, such as "3 of 41 suites ran". A subset is judged on what ran.
- A metric was not reported: a metric that normally reports has no value in this Result (see below).
The verdict, the Insights statistics and the CI check all ignore partial: a partial ✓ Healthy Result counts as Healthy.
A subset is measured against the Space's recent Results that weren't subsets. If smaller runs become a Space's normal, that reference moves and they stop reading partial. Sandbox, tag and aggregation-only Spaces are never subsets. In a Space with Manual testing, automated Results are measured by what each one ran, against earlier automated Results; a Manual session is never a subset.
CI Build Status and the CI Check
Testspace records your CI build status and shows it with each Result, but doesn't use it to judge anything. For Connected Projects, the check Testspace posts on your commit follows the verdict:
| Verdict | CI check |
|---|---|
| ✓ Healthy (partial or not) | success |
| ✕ Unhealthy (partial or not) | failure |
| ! Invalid | failure |
Metrics That Stop Reporting
A metric that has reported before but has no value in a Result is shown as not reported, and the Result is partial.
| Results without a value | What happens |
|---|---|
| 1 to 5 | not reported; the Result is partial |
| from the 6th | the metric's badge is hidden from lists until the metric reports again; the metric keeps its graph and still collects data, and its Metrics tab graph notes it |
| 15 Results and 30 days | metrics Testspace created (code coverage, static analysis) that you haven't edited are deleted, with their graph and history; Activity shows "Removed Metric …, by Testspace" |
- Subsets don't count toward these numbers. Neither do Invalid Results or pushes that reported an error, up to 5 in a row.
- Metrics you created, or Testspace-created metrics you have edited and saved, are never deleted; they stay hidden until data returns.
- A renamed coverage Suite is treated the same way: its old metrics read not reported, then hide. Its new metrics take the old ones' threshold (Code Coverage).
- A deleted metric's badge stops working everywhere, including a README badge or link to it. If the metric is published again, Testspace recreates it, without its old history.
- A gray NA is a metric with no value that wasn't expected, such as a new metric that hasn't reported yet. It isn't judged and doesn't make the Result partial.
Failure Tracking
How Testspace tracks a failing test from Result to Result (when it is resolved, when a test that stops running is retired, and how Manual testing differs) is on Failures.