Execution Status Frame

A run has a structure: a test runs its cycles, a cycle runs its cases, a case runs its case cycles, and a case cycle runs flows and steps. The tree reports the outcome of that structure once the pieces finish, and the log reports what happens inside them. The Execution Status frame, below the log in the right pane, reports where in the structure the run is right now: which stage is active at each level, how far along it is, and how long the whole run is expected to take. It is the frame to watch during a long run.

To see every level at once, the run has to have every level. The Remote Thread Example test from the Endpoints group, met in Handling Asynchronous Steps, is a good choice: its case runs one flow of three steps, and the middle step waits for a configurable number of seconds, long enough to watch. Select it, set Wait Sec to 20 and Cycles to 2, so that the test cycle level has something to count as well, and keep the endpoint at localhost:

Stages and progress

Click Update Plan, then Run, and look at the frame while the wait step is under way:

Each row is one level of the run, nested like the results tree, and Stage Name names the stage that is active at that level:

• the test, counting its cycles: the first of two;
• the test cycle, counting the same;
• the case with its parameters, the only case this test has;
• the case cycle, one per case here;
• the flow the case cycle is running, Remote Thread;
• the step the flow is running, Wait, with the endpoint it runs on after the @. A step is the unit that executes on an endpoint, which is why the endpoint is named on this row and no other.

Current and Total count that level's items, and the Progress bar shows the completed share. The flow and step rows have no counters: they are what is running, not a count of something. The header carries the ETA, the estimated time to the end of the run, computed from how long the completed items took, and a spinner that turns while the status is running. On a first cycle there is nothing to compute from yet, so the ETA shows a dash.

The frame follows the run as it moves. When the second cycle reaches its wait step, the test and cycle rows have advanced to two of two with their bars at half, the case rows have started over, and the ETA has a figure, taken from the first cycle:

The test's bar is purple, the colour of a running status, and it fills as cycles complete. The estimate settles after the first few items and is the figure to glance at when deciding whether to wait.

The frame reads together with the log above it. While the step row says Wait, the log's last line is the step's own message about waiting, and the endpoint card at the bottom left shows the endpoint as Busy:

The finished run

When the run ends, the frame keeps the last state of every level and takes the final status colour:

The test's bar is full and green, the frame's outline matches, and the rows show the last stage that ran at each level: the second cycle, its case, and the flow's last step, Read Maze Thread. A failed run leaves the same picture in red, with the rows pointing at where it stopped. The frame belongs to the selected run: selecting another run in the tree shows that run's state, and starting a new run resets it.

The collapsed frame

Click the arrow before the Execution Status title to fold the frame into one line:

The collapsed frame keeps the name of the running test and a single progress bar for it, so the run can still be followed while the log takes the rest of the pane:

Click the arrow again to expand the stages.

This completes the tutorial's tour of the controller's frames. The tree, the run panel, the log, the status frame and the Details dialog are the same views an AI agent works from when it runs the suites for you, through the files the controller exports rather than through the window; the AI Agents section picks the story up from there. For complete suites on real setups, a network and a hardware one, see the Examples.