Testing: Active Test
Overview
A test must already be running before this page shows anything, started from Run Test with a configured load profile or triggered by a schedule on Schedules. The agents configured for the run provide the capacity shown in the Agents card. Once you stop the test, or it ends on its own, the results leave this live view and move into the Test Runs history for later analysis.
This page lets you read the test's status and timing, watch KPI counters and throughput charts update in real time, open panels such as Login Outcome, Application Metrics, and Measurement Metrics, trace individual activity in the Live Event Stream, and act on the run with Pause, Resume, or Stop.
Key concepts
- SignalR - The live connection the page uses to stream metrics and events from the server in real time, without you needing to refresh the page.
- ramping up - The phase at the start of a run when virtual users log on gradually until the test reaches its configured concurrency. KPI counters and charts stay near zero until this phase progresses. (related: Testing: Load Profiles)
- Txn Success - The percentage of scripted transactions, business actions validated for correctness, that complete successfully. This is distinct from the raw connection errors shown in Errors/s. (related: Configuration: Testing: Scenarios)
- Agents - The units of load-generation capacity, provided by configured agent machines, currently assigned to and driving this run. (related: Configuration: LoadGen Agents: Full Agents)
When to use it
- To confirm a test you just started is actually ramping up users and generating load.
- To watch KPI counters such as Successful, Failed, and Errors/s for early signs the system under test is struggling.
- To decide whether to pause, resume, or stop a run based on live throughput and error trends.
- To trace a specific user's or action's activity in the Live Event Stream while a problem is happening.
- To confirm agent capacity shown in the Agents card is actually attached to the run you expected.
Before you begin
- A test must be currently running or starting: launch one from Testing: Run Test with a configured load profile, or let a configured schedule on Testing: Schedules trigger it.
- The load profile that defines what the test executes must already exist, created on Testing: Load Profiles.
- Agents must be configured and online so the run has capacity to draw on, set up on Configuration: LoadGen Agents: Full Agents.
- You need permission to view the Testing module to open this page.
Common workflows
Monitor a running test
Persona: Load-test operator
- Open Active Test while a test is running.
- Expect: The cockpit shows the test name, Running status, elapsed time, and KPI cards; while users are still ramping up, counters read zero and the charts show Collecting data...
- Watch the KPI cards and the Concurrency Trend and Event Throughput charts as the run progresses.
- Expect: Successful, Failed, Errors/s, and Events/s begin updating in real time, and the charts start plotting moving lines instead of showing Collecting data...
- Check SignalR: Connected and Last update periodically.
- Expect: Last update stays close to a few seconds, confirming the figures on screen are current rather than stale.
Verify: KPI counters and charts are actively changing rather than frozen at zero, and SignalR: Connected with a recent Last update confirms you are watching live data.
Pause and resume a test
Persona: Load-test operator
- Select Pause.
- Expect: Load generation stops, KPI counters and charts stop advancing, and the button changes to Resume.
- Select Resume.
- Expect: Load generation continues, KPI counters and charts start advancing again, and the button reverts to Pause.
Verify: The status pill still shows Running and the KPI counters resume climbing, confirming the test continued rather than ended.
Stop a test
Persona: Load-test operator
- Select Stop.
- Expect: A confirmation dialog, Stop Test, appears asking Are you sure you want to stop this test? This action cannot be undone.
- Confirm the stop.
- Expect: The test ends, the Pause and Stop controls disappear, and the run is no longer generating load.
Verify: The test no longer shows as Running on this page, and its completed data is available from Testing: Test Runs.
Steps
Step 1
Open the Active Test page while a test is running.
Expected result: The live cockpit loads showing the test name Test: LoadGen Demo, a Running status pill, an elapsed timer, and SignalR: Connected; because the run has just started ramping up users, the KPI cards read zero or a dash and both the Concurrency Trend and Event Throughput panels show Collecting data...
This is the starting view for checking a test's live progress before its metrics begin populating.
- The Running status pill confirms the test is active.
- SignalR: Connected confirms the live data feed is working.
-
Pause and Stop are available in the header while the test runs.
Field reference
| Field | Purpose | Format / Validation | Required | Default |
|---|---|---|---|---|
| Auto-scroll | Controls whether the Live Event Stream panel automatically scrolls to show each new event as it arrives. Turn it on to always see the latest activity; turn it off to keep reading an earlier event without the view jumping away from it. | on/off toggle | No | - |
| Search user or agent... | Filters the User Fleet list to the virtual users or agent machines whose name matches the text you type, so you can find one entry quickly among many. | free text search box | No | - |
| Filter action... | Narrows the Live Event Stream to events whose action matches the typed text, useful for isolating a specific transaction or step when tracing an error. | free text box | No | - |
| Filter user... | Narrows the Live Event Stream to events from one virtual user, useful when following a single user's session through the test. | free text box | No | - |
Column reference
| Column | What it shows | How to read it |
|---|---|---|
| Test status | Reflects the current lifecycle state of this test execution. | Running means load is actively being generated; use Pause or Stop to change this state. Other values correspond to a paused, stopped, or failed run. |
| elapsed | Shows how long the test has been running. | The annotation runs until manually stopped means this run has no fixed duration configured, so it continues until you select Stop; a test with a defined duration would instead count down toward an endpoint. |
| SignalR connection status | Reports the health of the live data feed this page depends on. | Connected with a green dot confirms the KPI counters and charts are receiving live updates; any other state means the values on screen may be stale. |
| Last update | Shows how long ago the page last received live data. | A value close to real time confirms the feed is current; a steadily growing age suggests the live connection has stalled even if SignalR still shows Connected. |
| In Progress | Count of virtual user sessions or transactions currently executing at this moment. | Rises as users ramp up and settles near the target concurrency once ramp-up completes; stays at zero while the test is still starting. |
| Successful | Count of completed transactions or sessions that finished without error so far in the run. | A steadily increasing count confirms the system under test is handling the current load correctly. |
| Failed | Count of completed transactions or sessions that ended in error so far in the run. | A rising count, especially one climbing faster than Successful, signals the system under test is struggling under the current load. |
| Errors/s | Rate of errors per second, a normalized view of Failed that stays comparable across different points in a long run regardless of how much time has elapsed. | A climbing Errors/s trend is a stronger warning than a high cumulative Failed count on a long run, since it reflects current conditions rather than history. |
| Events/s | Rate of logged events, user actions and steps, generated per second across all users: a general throughput indicator for the run. | Compare against the throughput you expect from the configured load profile; a value far below expectation suggests users are not reaching target concurrency or are blocked. |
| Txn Success | Percentage of scripted transactions completing successfully, distinct from the raw error rate shown in Errors/s. | Shows a dash until enough transactions have completed to calculate a rate; once populated, a falling percentage points to functional problems rather than connection-level errors. |
| Remaining | Indicates how much of the test is left, for example time or user count, when the run has a defined end condition. | Shows a dash for a test configured to run until manually stopped, since there is no fixed endpoint to count down to. |
| Agents | Number of agent slots currently assigned to and generating load for this run. | 0 slots briefly during startup is expected; a count stuck at 0 once the test should be running indicates no agents picked up the run, check Configuration: LoadGen Agents: Full Agents for availability. |
Controls reference
Pause
Temporarily halts load generation without ending the test.
When to use: Use it when you want to freeze the run's current state, for example to let the system under test recover or to inspect it without losing the run.
Side effects: Mutating: new load stops being generated, KPI counters and charts stop advancing, and the button changes to Resume.
Resume
Continues a paused test from where it left off.
When to use: Use it after Pause, once you are ready to continue generating load.
Side effects: Mutating: load generation resumes, KPI counters and charts start advancing again, and the button reverts to Pause.
Stop
Ends the test run.
When to use: Use it once you have enough data, the run is misbehaving, or you need to free up agents for another test.
Side effects: Mutating and irreversible: a confirmation dialog (Stop Test, Are you sure you want to stop this test? This action cannot be undone.) appears first; confirming ends load generation, moves the run toward its final results, and the Pause and Stop controls disappear.
Show Details
Expands the full error or initialization message when a test fails to start.
When to use: Use it when the page shows Test failed to start or Initialization error instead of the live cockpit, to see the specific reason.
Side effects: Read-only: reveals additional text and does not change the test.
Clear filter
Resets an active search or filter in a list panel back to showing every row.
When to use: Use it after filtering the User Fleet or Live Event Stream and you want to see everything again.
Side effects: Read-only: changes only what is displayed, not the underlying test.
Pin event
Keeps a specific Live Event Stream row visible and marked while new events continue to stream in below it.
When to use: Use it to keep an important error or milestone event in view while Auto-scroll keeps adding newer events.
Side effects: Read-only: local display state only, does not affect the test.
Back to Run
Returns you to the page where you configured and launched the test.
When to use: Use it after a test fails to start, to adjust settings and try again.
Side effects: Read-only navigation.
Start New Test
Takes you to start a new test run.
When to use: Use it when this page shows No Test Running and you want to launch one.
Side effects: Read-only navigation.
Open Test History
Takes you to the history of past test runs.
When to use: Use it when no test is currently active and you want to review previous runs instead.
Side effects: Read-only navigation.
Open in Results
Takes you directly to the detailed results for a run that has finished.
When to use: Use it once a test completes, to review full results instead of the live cockpit.
Side effects: Read-only navigation.
Verify the result
When the primary task succeeds, the Active Test page shows the test's name with a Running status pill, a SignalR: Connected indicator, and an elapsed timer counting up. KPI cards for In Progress, Successful, Failed, Errors/s, and Events/s update continuously instead of staying at zero, and the Concurrency Trend and Event Throughput charts plot moving lines instead of showing Collecting data... The Agents card shows a nonzero slot count once agents have picked up the run. If values stay at zero or the charts never leave Collecting data..., check the SignalR connection status first, then confirm agents are online.
Troubleshooting
- Symptom: The KPI cards stay at 0 and the charts stay on Collecting data... well after the Starting users: ramping up... message. Fix: Confirm SignalR: Connected is still shown; if the connection has dropped, refresh the page and check that agents are online on Configuration: LoadGen Agents: Full Agents.
- Symptom: A banner reports that polling has failed after several attempts. Fix: Treat this as a lost live-data connection rather than a stopped test; refresh the page to re-establish the feed before deciding whether to stop the run.
- Symptom: Selecting Pause reports that pausing the test failed. Fix: The test keeps running as if nothing happened; wait a moment and try Pause again, or use Stop if the run needs to end regardless.
- Symptom: Selecting Stop reports that stopping the test failed. Fix: Assume the test may still be running, reload the page to confirm its state, and retry Stop if it still shows Running.
- Symptom: The page opens straight to a no-test-running state right after you launched one. Fix: Confirm the test actually started from Testing: Run Test; if it failed during initialization, this page instead shows Test failed to start with a reason.
- Symptom: The page shows Test failed to start or Initialization error instead of the live cockpit. Fix: Select Show Details to read the specific reason, then use Back to Run to return to Testing: Run Test and correct the configuration.
Related articles
- Prerequisite: Testing: Run Test - a test run must be launched here before Active Test has anything to show
- Prerequisite: Testing: Load Profiles - defines the scenario, user count, and ramp pattern the active run executes
- Prerequisite: Testing: Schedules - scheduled runs also populate this live view automatically
- Next step: Testing: Test Runs - a stopped or finished run's full results move here for later analysis
- See also: Configuration: LoadGen Agents: Full Agents - provides the agent capacity reflected in the Agents KPI card