Scanning
A scan is how data gets from an integration into the asset inventory. Connecting an integration on its own does nothing: it only collects when a scan template that uses it is run, by hand or on a schedule.
Four things are involved:
| Thing | What it is | Where |
|---|---|---|
| Integration instance | A configured connection to a source, assigned to a sensor | Settings → Integrations |
| Scan template | One integration instance, that integration's scan settings, and optional rules applied to what it finds | Scans → Templates |
| Schedule | A template plus a cron expression that runs it on a repeating cadence | Scans → Scheduled |
| Scan run | One execution of a template, with its stages, its findings and the changes it proposes | Scans → History |
One integration instance can back several templates (for example one that creates assets automatically and one that holds new assets for approval), and one template can be run by several schedules.
The pages in this section cover each part in detail:
- Scan templates: creating a template, its settings, rules, and templates that scan an uploaded file.
- Schedules: cron syntax, the UTC clock, pausing and what happens when a run overlaps the next slot.
- Reviewing a scan: the run page, resolving conflicts, discarding changes and approving a run.
Permissions
Each part has its own tenant permission (see Roles and permissions):
| Action | Permission |
|---|---|
| View, create, edit or delete scan templates | Scans (Read / Write) |
| View, create, edit, pause or delete schedules | Scan schedules (Read / Write) |
| Start a scan (Run now, Run again) or delete a run | Task results (Write) |
| Resolve conflicts, discard changes, approve a run | Scan results (Write) |
Running a scan now
Every template can be started by hand, independently of any schedule:
- On Scans → Templates, use the row's Run action, or tick several rows and choose Run scans.
- On a template's own page, use Run now.
- On Scans → History or a run's page, use Run again to start the same template once more.
The scan is queued at once and appears in Scans → History. A few rules apply:
- A template whose integration instance is assigned to a disabled sensor cannot be started; the console shows an error instead.
- Starting a template that already has a run in progress queues a second run. The console does not wait for, or cancel, the first one.
- A template whose integration is set up to analyse an uploaded file (an isolated integration) shows Run with a file instead. It asks for the file before it starts, and is skipped when you run several templates at once. See Templates that scan a file.
The scan list
Scans → History lists every run in the tenant, newest first. Each row shows the template it ran (Scan), its State, when it Started, its Duration (still counting for a live run), and how many Changes and Findings it produced. Integration, Finished and Run id can be added from the Columns menu.
Above the table you can search by template name, and filter by State and by Template. Search, filters and paging are kept in the page address, so a filtered view can be shared as a link. The list refreshes itself while any run on the page is still moving.
Row actions are Run again and Delete run; ticking rows offers Delete runs. Deleting a run removes it and its findings from history. Assets it has already created or updated stay as they are: deleting a run does not roll anything back.
Stages
Every scan walks the same nine stages in order. The first three happen on the sensor; the rest happen in Velum Edge after the upload. A run's page shows them as a pipeline, and any stage that wrote a log can be opened to read it.
| Stage (as the console shows it) | Runs on | What happens |
|---|---|---|
| Queued | Sensor | The run is waiting for its sensor to collect it |
| Collecting | Sensor | The sensor reads from the source, or scans the network |
| Uploading | Sensor | The sensor sends what it collected back |
| Verification | Velum Edge | The upload is unpacked and split into one raw finding per record |
| Cleaning | Velum Edge | Housekeeping between stages; most integrations pass straight through |
| Identification | Velum Edge | Each record is turned into a change entry and matched against existing assets; the template's rules run here. A run stops here when changes need a decision |
| Validation | Velum Edge | The template's action settings are checked; a run stops here when changes need approval |
| Syncing | Velum Edge | The changes are written into the inventory |
| Finished | Velum Edge | Nothing left to do |
A run that stays at Queued or fails at Collecting points at the sensor side: the sensor is offline, cannot reach the source, or its credentials were refused. A run that fails after Uploading points at processing in Velum Edge, and the sensor is fine. Open the failing stage's log on the run's page to see why.
Statuses
Alongside its stage, every run has a status. The console combines the two into one state, shown in the State column and at the top of a run's page:
| State | Status behind it | Meaning |
|---|---|---|
| Queued | Running or pending, at the Queued stage | Waiting for a sensor to pick the scan up |
| Running | Running, pending or completed, at any later stage | Collecting or processing; nothing for you to do |
| Needs review | Conflict | Some changes matched more than one asset, or need a decision. The run waits at Identification until you resolve them |
| Awaiting approval | Verification | The changes are resolved and held for approval because of the template's action settings. The run waits at Validation |
| Completed | Completed, at the Finished stage | The changes are in the inventory |
| Failed | Error | The run stopped at the stage shown; its log says why |
The State filter on the scan list offers In progress, Needs review, Awaiting approval, Completed and Failed. In progress covers both queued and running scans.
A run in Needs review or Awaiting approval does not move on by itself. See Reviewing a scan for what to do.
Scan history retention
Velum Edge keeps the newest 50 scans per tenant by default; your deployment can set a different limit per tenant. When a new scan starts, and again during hourly housekeeping, scans beyond the newest 50 are deleted together with their findings, change entries and stored files. Assets they created or updated are not affected.
A scan beyond the limit is deleted when nothing more will come of it:
- it has finished or failed;
- it is waiting in Needs review or Awaiting approval;
- it is still queued at the sensor;
- it has been running for 3 days (by default) without any change, which counts as stuck.
A scan beyond the limit that is still running and still making progress is kept until it finishes or becomes stuck.
A scan waiting for you to resolve conflicts or approve its changes is deleted like any other once 50 newer scans exist. On a tenant with many frequent schedules, review scans promptly or their proposed changes are lost.
See also
- Integrations: the sources a template scans with, and the scan settings they share
- Sensors: the component that runs every scan
- Assets: where synced changes land