Status holds every action on the deployment in its action bar. Each action asks for confirmation first, runs one job at a time, and can be cancelled while it runs.
| The job | The toast reads |
|---|---|
| Succeeded | DataMind OS started, DataMind OS stopped, DataMind OS restarted or DataMind OS terminated |
| Failed | Couldn’t start DataMind OS, Couldn’t stop DataMind OS, Couldn’t restart DataMind OS or Couldn’t terminate DataMind OS |
A failed action adds a job logs link to the toast; it opens the full output of the job. Troubleshooting lists each message with its cause and fix.
The action bar, Cancel and the restart buttons appear only for admins. An admin sets roles on Users.
| Action | Asks first | What it does |
|---|---|---|
| Update | Update DataMind OS? | Pulls the latest images and recreates the affected services, so running services restart briefly. Steps, phases and recovery: Updating |
| Start | Start DataMind OS? | Brings all stopped services back up, and recreates any service whose saved settings changed |
| Stop | Stop DataMind OS? | Stops every service gracefully. DataMind OS stays offline until you start it again |
| Restart | Restart DataMind OS? | Restarts the services in place. Expect brief downtime; see the configuration note below |
| Terminate | Terminate DataMind OS? | Force-stops all services immediately (SIGKILL, no graceful shutdown). |
Configuration changes are NOT applied, and it tells you to save your configs and use Apply changes for that; see Configuration.{service} restarted or {N} services restarted. One-shot services have no restart button. See Service status.The Installer runs one operation at a time. While a job runs, every control in the bar is disabled
and its tooltip gives the reason: Available when the <label> finishes. Another action is refused
with A <type> job is already running (started <time>). While the Installer
updates itself, every control reads Available when the Installer update finishes.
DataMind OS is already running. When its images are missing, the tooltip counts them (N images missing) and says to run Update first.DataMind OS is not running.DataMind OS is already up to date. The check icon beside it shows when the last check ran: Check for updates · last checked <time>.An admin account is required to start DataMind OS.While a job runs, Cancel is offered. Jobs that recreate services through Compose ask for
confirmation first: Cancel this job?. Other jobs cancel immediately. Cancelling while the images
are still being pulled changes nothing.
The confirmation notes that Docker compose is already recreating services and that
Some containers may have been recreated: if that happened, finish the deployment with Start, or
remove the extra containers with Stop. A cancelled job ends with the toast Cancelled; if the
cancel itself fails, the toast reads Couldn’t cancel job.
No. Restart brings the services back with the settings they were created with, so saved changes stay inactive. Click Apply changes on Configuration to make them live.
Stop shuts every service down gracefully. Terminate force-stops them immediately (SIGKILL, no graceful shutdown), and in-flight writes can be lost. Use it only for a service that does not answer Stop.
Most often because DataMind OS already runs: DataMind OS is already running. Otherwise its images
are missing: the tooltip counts them (N images missing) and says to run Update first. Before
the first install, Start stays disabled until an admin account exists.
The Installer runs one operation at a time. Every control is disabled, with the tooltip
Available when the <label> finishes, and another action is refused with
A <type> job is already running (started <time>). While the Installer
updates itself, the tooltip reads Available when the Installer update finishes.
Yes. Cancel is offered while a job runs. Jobs that recreate services through Compose ask
Cancel this job? first; other jobs cancel immediately. Cancelling during the image pull changes
nothing; once Compose has begun, some containers may be part-way: finish with Start, or remove
them with Stop.
Yes, possibly. Terminate force-stops every service immediately (SIGKILL, no graceful shutdown), so in-flight writes can be lost. Use Stop first, and terminate only a service that ignores it.
Start brings all stopped services back up and recreates any service whose saved settings changed. If you expected a plain resume, check Configuration for a recent save.
Yes. Use the restart button on a group or a service in the Services panel; those restarts work
in place and do not apply configuration changes. They report {service} restarted or
{N} services restarted. One-shot services have no restart button. See
Service status.
No. The job runs on the host. If the page loses its connection, it reconnects on its own, and the running job appears again when you come back.