Deployments
The Deployments page shows every deployment job across your environment. Use it to monitor progress, troubleshoot failures, and retry or cancel jobs.
In short
- Each row is the latest attempt of a deployment, so a failed attempt that was retried and then succeeded counts as succeeded.
- Filters cover every deployment in the tenant and are kept in the page address, so you can bookmark or share a filtered view.
- Pckgr retries a failed job on its own up to its attempt limit; after that you can still retry it, which starts the count again.
- The logs on a deployment's detail page hold the full output from the installer and the agent's execution steps.
The deployments list

| Column | What it shows |
|---|---|
| Created | When the job was created (shown as relative time) |
| Device | Target device hostname |
| App | Application name |
| Intent | Install, Update, Uninstall, or Update Only |
| Status | Current status of the job |
| Attempt | Current attempt number (for example, “1/3” for the first of three allowed attempts) |
| Failure Type | Reason for failure, if applicable |
| Duration | How long the job took or has been running |
Job statuses
| Status | Meaning |
|---|---|
| Pending | Job is queued, waiting for the device to pick it up. A job for an app that runs as the signed-in user also waits here while nobody is signed in. On Windows that includes Autopilot pre-provisioning and setup, because the temporary account Windows signs in during setup does not count as a user. |
| InProgress | Job is currently executing on the device |
| Succeeded | Job completed successfully |
| Failed | Job encountered an error during execution |
| Cancelled | Job was cancelled by an administrator |
| Skipped | Job was skipped (for example, the app was already installed) |
| Abandoned | Device stopped responding during execution (missed heartbeat) |
| Deferred | Waiting to run again. The user postponed the install, nobody answered the install prompt before it closed, or no one was signed in for an install that runs as the signed-in user: each way it is offered again 12 hours later and uses no attempt. An update also waits here until the app's Update Delay has passed. |
Filter deployments
- Status
- Job status, for example only Failed jobs.
- Intent
- Deployment intent, or Any install for Install, Update and Update Only together.
- Device
- A specific device.
- App
- A specific application.
- Search
- Device hostname or app name.
Filters cover every deployment in the tenant, not only the most recent ones, and the list shows 50 at a time with Previous and Next below it. The filters are kept in the page address, so you can bookmark a filtered view or send it to a colleague.
Clicking a count on an app's page opens this list filtered to the same devices, with a Latest job per device filter that matches how the app page counts. Remove it to see every deployment of that app.
Click Clear to reset all filters.
Custom uninstall jobs
When you remove an app from a device's Detected Applications list, the resulting job appears here with a Custom uninstall badge next to the app name.
These jobs aren't tied to a catalog app; they target whatever was installed on the device, so they don't follow assignments or retry rules. Each is a one-shot action you initiated from the device detail page.
Retry history
Deployments are grouped by their root job. If a job has been retried, click the expand arrow to see all previous attempts. The latest attempt is shown at the top, with earlier attempts listed below.
Note
Deployment details
Click any deployment row to open its detail page.
- Job information
- Job ID, device, app, intent, status, timestamps, duration, exit code, and failure details.
- Logs
- Browse log files collected from the device, view individual log file contents, copy the one you are viewing to the clipboard, or download the complete log archive. Logs include standard output and error from the installer, agent execution logs, and PSADT toolkit logs.
Cancel or retry a job
Cancel a job
Open the detail page of a pending, deferred or in-progress job.
Click Cancel.
A pending or deferred job is cancelled straight away. For an in-progress job, a cancellation request is sent and the job stops on its next status check.
To cancel several jobs at once, select them in the list and use Cancel selected (see Bulk operations below).
Retry a job
Click Retry on a failed, abandoned, or cancelled job. This creates a new job attempt for the same device and app. On a job that has used every attempt (for example, attempt 3 of 3), Retry starts again at attempt 1, and Pckgr's automatic retries pick up from there.
Bulk operations
Select multiple deployments using the checkboxes, then use the action bar:
- Retry failed
- Create new attempts for the selected failed, abandoned, or cancelled jobs. A job that has used every attempt starts again at attempt 1.
- Cancel selected
- Cancel the selected pending, deferred, or in-progress jobs. Pending and deferred jobs are cancelled straight away; in-progress jobs are asked to stop.
Jobs that can't be retried or cancelled are skipped, and the result message tells you how many were.
Failure types
When a job fails, the Failure Type column gives more context:
| Failure type | Meaning |
|---|---|
| Download Failed | The package could not be downloaded to the device |
| Extraction Failed | The package downloaded but could not be unpacked |
| Requirement Failed | A requirement check did not pass (for example, an Update Only job on a device without the app), or an app this one depends on failed |
| Detection Failed | The detection check could not run, or the installer finished but the app was not detected afterwards. Some installers finish their work in the background after they return; if the device detects the app at a check-in within a day, the job changes to Succeeded |
| Execution Failed | The installer or script ran and reported an error |
| Timeout | The job ran past its time limit (60 minutes for most installs) |
| Log Upload Failed | The job’s logs could not be uploaded from the device |
| Cancelled | The job was cancelled, by an administrator or because an app it depends on was cancelled |
| Device Offline | The device stopped responding while the job was running, or lost its hold on the job |
| Deferred | The user postponed the installation, nobody answered the install prompt before it closed, or no one was signed in for a signed-in-user install. It is offered again 12 hours later |
| Unknown | The cause could not be determined; check the logs |
Tips
- Check the Dashboard for a quick view of deployment trends and top failures.
- Use filters to find specific failures. For example, filter by Status = Failed and a specific App to investigate recurring issues.
- Always check the logs when troubleshooting: they contain the full output from the installer and the agent's execution steps.