Preview-first Dataverse bulk upsert utility inside VS Code.
DV Bulk Upsert Runner is a focused DV ForgeLab utility for importing CSV, JSON, or .dvbur.json packages, validating them, previewing single-entity upsert operations, applying them deliberately, and reviewing execution outcomes.
It is intentionally about data application, not migration.
DV Bulk Upsert Runner applies staged single-entity data rows. It does not perform ETL, scheduled sync, relationship graph migration, attachment migration, or automatic data cleansing.
Highlights
- CSV import
- JSON import
.dvbur.jsonpackage support- Single-entity bulk upsert workflow
- Primary ID support
- Alternate-key support
- Metadata-backed key selection
- Create / Update classification
- Preview-first execution
- Batch execution progress tracking
- Failure grouping and review
- Failure export
- Failure requeue
- Environment-aware execution
- Shared DV ForgeLab environment settings
- Direct DV ForgeLab feedback integration
Preview-First Workflow
Connect
↓
Import CSV / JSON / DVBUR Package
↓
Validate
↓
Preview Rows
↓
Check Creates / Updates
↓
Apply Upserts
↓
Review Results
↓
Export / Requeue Failures
Screenshots
Import & Preview

Import CSV, JSON, or DVBUR packages and preview staged rows before Dataverse is modified.
Create / Update Classification

DVBUR can classify staged rows as creates or updates before execution.
Metadata-backed key selection helps identify valid primary IDs and active alternate keys.
Applying Upserts

Execution is performed in batches with live progress, throughput, elapsed time, and estimated completion tracking.
Preview Review Surface

Review classified creates and updates before applying changes.
DVBUR follows a preview-first workflow and never modifies Dataverse without explicit user action.
Failure Analysis

Execution failures are grouped and summarised.
DVBUR surfaces:
- failure category
- affected column
- invalid value
- expected type
- suggested corrective action
- raw Dataverse technical details
Failures can be exported or requeued for later correction and replay.
Supported Scope
Supported
- Single entity imports
- CSV records
- JSON records
- DVBUR package records
- Primary ID upserts
- Alternate-key upserts
- Batch execution
- Failure export
- Failure requeue
- Create / Update classification
Not Supported
- ETL pipelines
- Scheduled synchronisation
- Relationship graph migration
- Attachment migration
- Data cleansing
- Transformation workflows
- Cross-entity dependency resolution
- Enterprise migration orchestration
Boundary
DV Bulk Upsert Runner is intentionally a runner, not a migration platform.
It does not:
- Perform ETL
- Schedule synchronisation
- Transform data
- Cleanse data
- Infer mappings
- Migrate relationship graphs
- Migrate files or attachments
- Automatically repair data quality issues
- Replace enterprise migration tooling
Shared DV ForgeLab Environment Settings
"dvForgeLab.environments": [
{
"name": "DEV",
"url": "https://org.crm6.dynamics.com",
"tenantId": "optional-tenant-id"
}
]
Command
DV Bulk Upsert Runner: Open Upsert Runner
Feedback
DV Bulk Upsert Runner includes direct integration with the DV ForgeLab feedback portal.
Share:
- Feature requests
- Bug reports
- Bulk upsert scenarios
- Workflow suggestions
- Product feedback
Feedback is routed through the shared DV ForgeLab feedback experience and automatically identifies DV Bulk Upsert Runner as the source product.
https://www.dvforgelab.com/feedback
Philosophy
DV Bulk Upsert Runner follows the DV ForgeLab preview-first invariant.
Rows are imported locally, validated, previewed, classified, and explicitly applied by the user.
Dataverse data is never changed without an explicit review and apply step.
Part of the DV ForgeLab Family
DV Bulk Upsert Runner is a focused Dataverse utility from DV ForgeLab.
For operational investigation, execution, runtime analysis, and cross-environment comparison, see DV Quick Run.
DV Bulk Upsert Runner follows the same principles:
- Preview-first
- Environment-aware
- Metadata-backed
- Explicit execution
- Calm operational UX
Built by DV ForgeLab.