You can reorganize later (Pro)

The safest hierarchy plan is one you can change. DefectDojo Pro lets you get scan data in on day one with a reasonable structure, then move Assets between Organizations, change their parents, or point Connector Records at different Assets as you learn more, without losing Findings, their history or their deduplication state.

This page lists, for each kind of change, what moves, what is recalculated, and what stays exactly as it was. If you are still deciding on a structure, start with Plan your hierarchy.

Moving an Asset to another Organization

Change the Organization on the Edit Asset form, use Bulk Edit on the All Assets list, or send the change through the API. The same rules apply whichever way the move is made, including moves made by the Rules Engine.

What happens

  • The Asset’s children move with it. Every Asset below it in the hierarchy moves to the new Organization too, so the tree stays whole.
  • A parent link left crossing Organizations is removed. Parent and child Assets always share an Organization, so when an Asset moves away from its parent, the link to that parent is removed and the Asset becomes a top-level Asset in its new Organization.
  • Connectors do not undo the move. A Connector that declares a parent in another Organization skips that link and logs it, rather than pulling the Asset back.
  • People are told. The owners and contacts of each Asset that lost its parent receive a notification naming the Assets and where they went. The removed links are recorded in the audit log.
  • Priority is recalculated in both Organizations. Revenue and User Records are weighed as a share of their Organization’s totals, so a move changes the picture for every Asset in the Organization it left and the one it joined. DefectDojo recalculates Finding Priority and Risk for both Organizations, once each, however many Assets moved. Where an Asset uses a Risk-based SLA Configuration, a Finding whose Risk changes gets the deadline for its new Risk level.

What changes and what stays

ChangesStays the same
The Organization of the Asset and of every Asset below itThe Asset itself: its ID, name and settings
The parent link, if the parent stays behindIts Engagements, Tests and Findings, with their notes, files and history
Who can see it through an Organization role: roles on the old Organization no longer reach it, roles on the new one doDeduplication state: duplicate links, and the hash codes reimport uses to match the next scan
Organization-filtered metrics, dashboards and reports pick it up under the new OrganizationRoles given on the Asset itself
Finding Priority and Risk, recalculated for both OrganizationsIts five priority fields, Prioritization Engine, SLA Configuration and tags
Its Connector Record mappings, CI import targets and Jira or integrator mappings, which all point at the Asset rather than its Organization
Its Dedupe Pool memberships

Because deduplication works inside each Asset, and a moved Asset keeps its own Findings, the next import or Connector sync into it matches what is already there. Nothing is re-imported and nothing is duplicated.

Permissions

Moving an Asset needs edit permission on it and permission to add Assets to the destination Organization. Because its children move too, you also need edit permission on every Asset that moves with it. If you lack it on any of them, the move is refused with a message naming the Asset, and nothing is changed.

Moving many Assets at once

Bulk Edit on the All Assets list moves the selected Assets together and can also set a parent (or remove it), the Asset type (when Asset types are turned on), Business Criticality, the SLA Configuration, the Prioritization Engine and tags. When you choose an Organization, Move Children Along decides what happens below each selected Asset:

  • On (the default): children move with their parent, as described above.
  • Off: children stay in the old Organization. Their links to the moved Asset are removed, and they become top-level Assets there.

The same update is available to scripts as POST /api/v2/assets/bulk_update/, which also accepts the three contacts (technical contact, team manager, product manager). It takes up to 500 Assets per request and is all or nothing: if any Asset in the batch cannot be changed (no permission, a parent in another Organization, a cycle, a child that cannot move along, or Findings still being recalculated from an earlier SLA Configuration or Prioritization Engine change), the whole request is refused with a list of every problem, and nothing is written. Split larger reorganizations into batches of 500.

Tags given in a bulk update are added to each Asset’s existing tags. Nothing is removed.

Changing an Asset’s parent

Use Change Parent, Add Child or Remove From Hierarchy on the Asset Hierarchy screen, the parent field on the Edit Asset form, or Bulk Edit.

  • The new parent must be in the same Organization. To put an Asset under a parent in another Organization, move it to that Organization first.
  • A change that would create a cycle (an Asset ending up below its own child) is refused.

Re-parenting changes how Assets roll up, and nothing else:

ChangesStays the same
The tree shown on the Asset Hierarchy screenThe Asset’s Organization, Findings and history
The indirect Finding counts of the old and new parents, which are always calculated from the current treeWho can see the Asset: a parent never grants or limits access
Metrics that include child AssetsDeduplication: Findings still deduplicate within their own Asset
Priority, Risk and SLAs: they come from the Asset’s own fields and its Organization

One exception to plan for. If you used Pool this asset and everything under it to put a subtree in a Dedupe Pool, those memberships were made from the tree as it stood then, and re-parenting does not update them. After a large re-parent, run Untoggle Subtree and pool the subtree again to bring the pool in line with the new tree.

Re-mapping a Connector Record

A Connector Record is what links an Asset in your scanning tool to an Asset in DefectDojo. When you point an already-mapped Record at a different Asset from Edit Record on the Manage Records page, DefectDojo asks what should happen to the Findings that Record has already imported. See Change the Mapping of a Record.

Move (the default)

The Record’s Tests move to the new Asset, under its Global Connectors Engagement, and keep their link to the Record. The next sync reimports into the same Tests, so it matches the Findings already there: no duplicate set, and no Findings closed and reopened.

  • Travels with the Tests, untouched: notes, files, history, Finding Groups, import history, and the hash codes deduplication and reimport use.
  • Carried across, because they belong to an Asset: endpoint and location statuses, Risk Acceptances, inherited tags, and SLA dates, which are recalculated for the new Asset’s SLA Configuration.
  • Recalculated: Finding Priority and Risk, and the grades of both Assets.
  • Released: a duplicate link whose original stays behind and would no longer be in the same deduplication scope.
  • Recorded: the moved Test gets a note naming the old and the new Asset.

The Connector Engagement left empty on the old Asset is removed, unless something a person added is attached to it.

Start fresh

The old Tests stay on the old Asset as history. Their open and risk-accepted Findings are closed with a note saying the Record was re-mapped, and the Record starts importing into new Tests under the new Asset.

Choose this when the old Findings no longer describe anything real, for example when the Record was mapped to the wrong system entirely.

Permissions

Re-mapping a Record needs permission to edit the Connector, and edit permission on both the old and the new Asset, so a re-map can never move Findings out of, or close Findings in, an Asset you cannot edit. A first mapping of a new Record has nothing to carry and asks no question.

Pipelines that still use the old names

CI pipelines and scripts import by name: product_type_name and product_name on /api/v2/import-scan/, /api/v2/reimport-scan/ and /api/v2/sbom-import/, or organization_name and asset_name on /api/v3/import. After you move or rename an Asset, or rename an Organization, a pipeline that still sends the old names keeps working:

  • A moved Asset. When the pipeline sends the Organization the Asset was in before the move, or a former name of the Organization it is in now, the import goes into the Asset. Before, the import was refused because the Asset was no longer in that Organization.
  • A renamed Asset. When no Asset has the name the pipeline sends and exactly one Asset had it before a rename, the import goes into that Asset instead of creating a new Asset under the old name.
  • A renamed Organization. When an import creates a new Asset (auto_create_context) in an Organization that has since been renamed, the Asset is created in that Organization instead of in a second Organization under the old name.

The response then includes name_resolution: the names the pipeline sent, the current names the import used, and a message saying what to change in the pipeline. Update the pipeline when you see it.

Some things stay exactly as they were:

  • An Organization the Asset was never in is still refused, with the same error as before.
  • A former name stops working after 180 days, or as soon as another Asset or Organization takes that name. Expired former names are deleted every night.
  • A name that two Assets had at different times resolves to neither of them.
  • Names are matched exactly, including upper and lower case.
  • Only a user who can view the Asset gets this. Anyone else gets the answer the import gave before, so an import never tells someone where an Asset they cannot see has gone.

Connectors, the Rules Engine, Smart Upload, cloud account onboarding and the GitLab project import also find an Asset or Organization by a former name, so a rename does not make them create a duplicate under the old name. Two exceptions keep that safe:

  • A Connector does not hand a former name to an Asset that another Record of the same Connector already maps. When the tool itself renamed that Record, a new Record that later takes the old name is a different thing, and it gets its own Asset.
  • The Rules Engine, Smart Upload, cloud account onboarding and the GitLab project import only use a former name of an Asset or Organization the person behind them can already see. For anyone else the name is free, as it was before.

To go back to the strict behavior, turn off Imports Follow Renames and Moves on the Feature Flags page. Former names are still recorded while it is off, so turning it on again finds them.

Before you start moving things

  • Check who will gain or lose access. Organization roles decide most visibility, so a move between Organizations is also an access change. Roles given on individual Assets travel with the Asset.
  • Expect a recalculation. After a large move, Priority and Risk update in the background for both Organizations. Lists sorted by Priority settle once it finishes.
  • Move whole trees where you can. Moving a parent takes its children along and keeps the tree intact. Moving children out one at a time removes parent links you may want to keep.
  • Batch big reorganizations. The bulk update takes 500 Assets per request and refuses a batch with any problem, so a refused batch never leaves a half-finished reorganization behind.

Next steps