Dynamics 365 F&SCM · Connectivity
D365 F&SCM Connectivity FAQs
Get trusted D365 F&SCM data without the bottlenecks
Zap's X++ extractor delivers Dynamics 365 F&SCM teams near-real-time data loads without throttling production. The FAQ below breaks down how it works, why OData/BYOD/Fabric exports fall short for the month-end close, and answers key questions about performance, delete capture, and roadmap.
| Question | Answer |
|---|---|
| How does Zap extract data from D365? | A signed deployable X++ package that runs as a batch job, pulling only delta records since last extraction using timestamp-based incremental processing. |
| Why not use OData, BYOD, or Synapse / Fabric? | OData is near-real-time but too slow for warehouse loads. Lake/Synapse/Fabric exports land every 5-15 min (typical) and can stretch to ~60 min on very large tables. Too slow for finance teams at month end. |
| Is the X++ extractor a workaround? | No. It follows the same deployable package model used by every Dynamics ISV solution (tax, EDI, etc.). We support both classic LCS deployments and Power Platform deployments using unified packages. |
| Why hasn't Microsoft built the same real-time feed? | Microsoft provides solutions that fit within their broader ecosystem of tools and product strategy. These tools are strong for the general use case, but not strong for real-time finance use cases. |
| Could Microsoft block this later? | Highly unlikely. Removing deployable packages would break hundreds of ISV solutions. API changes carry 12-24 months' notice. |
| Will it impact D365 performance? | Queries run NOLOCK as low-priority batch jobs. If D365 auto-kills the job to protect performance, Zap resumes from the last processed partition, and the UI threads stay unaffected. |
| How are deletes handled? | We copy posted rows only. Hard deletes are reconciled via a scheduled partition reload (weekly default, nightly optional). |
| Query pattern & indexing? | Single-table reads or header/line joins on timestamp. Ad-hoc SQL not supported. All modeling is performed in the warehouse. Zap flags a few one-time indexes for high-volume tables. |
| Can we still use Fabric or Synapse? | Yes. Zap can source from Fabric or Synapse as an alternative when needed. We default to X++ to significantly reduce latency and setup. |
| Roadmap? | Upcoming release adds entity-key back-mapping for quicker delete detection. |