Run
In the automation group, select the automations and click Run. The Run Automation wizard walks you through the run in steps; each step shows a check mark when it is ready, or what is missing. The footer summarizes the run and lists what is still missing under Missing:.
The button is disabled when the cycle or the group is closed, and for anyone who cannot run in the Automation Center.
What will run
The selected automations, with one dataset each. Here you can adjust the parameters for this run only:
- By dataset: open a dataset and change the values. This is also where you deselect a dataset that should not run.
- By parameter: change a parameter in all the selected datasets at once.
Only fixed values and lists can be changed here; encrypted values and references are read-only, and outputs are not shown. A changed value is highlighted.
The adjustment applies only to this run (and to the scheduled run, when it goes out). The dataset does not change. To change the dataset, edit it in Automations.
Where it runs
Choose a Runner or a Cluster.
- Runners are shown as Free, Running, or Offline, with free ones first. The first free runner is already selected. Show offline includes the ones that are turned off.
- An offline runner can be used to schedule, but not to run now.
- When an automation can only run on certain runners, the list shows only the allowed ones. A private runner only shows up for its owner.
Device and app
With a mobile automation, the step becomes Where it runs and device:
- Device on the runner: a device connected to the runner, or Any device (the first one connected at that moment). On a cluster, any device on the runner that picks up the run will do. Search again refreshes the list.
- BrowserStack: a BrowserStack device, with search and an Android and iOS filter. Only available when the BrowserStack integration is set up.
- App: the app and the version, Always the latest or A specific version. With the latest, scheduled and recurring runs pick up whatever version arrives later. Apps are kept in Resources > Mobile Apps.
- Record video: records the run on the device.
When
| Option | What it does |
|---|---|
| Now | The run goes into the runner's queue right away. |
| Schedule | Once, on the chosen date and time (in your time zone). |
| Recurring | Repeats on its own: every N minutes or hours, every day, on days of the week, or on a day of the month. See Schedule. |
Schedule and Recurring depend on the plan; without the feature, the option shows a lock.
Code and environment
- Repositories: the branch for each system. The repository's default branch is already selected; a different branch applies only to this run, for example to test a working branch before the merge.
- Environment: the run's environment, or Default of each dataset. When the run uses a production environment, the wizard warns you. See Environments.
More options
- Receive notification after execution: an email when the last run of the batch finishes, with the score and the link to each failure.
- Re-run on execution failure: a new attempt of the run that failed, right away or after a delay (seconds, minutes, or hours), with the same runner, the same parameters, the same branches, and the same environment. It is a single attempt, and it does not apply to a stopped run.
Send
The final button changes with the chosen option: Run, Schedule, or Save as Recurrence. It is only enabled when nothing is pending. Once the run is sent, the wizard shows where the runs went and closes on its own.
The first run of a repository on a runner downloads the code and prepares the dependencies before the first step, and can take a few minutes.
Permissions
Running, scheduling, and creating a recurring run require run permission in the Automation Center. The wizard also reads runners, clusters, systems, dataset parameters, and environments: the profile needs to see those screens for each step to show up complete.