Transfer-Inbound Sync to Kingdee Other-Outbound (Negative): Inventory Sync Strategy Tutorial
What This Strategy Solves
In supply chain integration scenarios for retail or manufacturing enterprises, inter-warehouse inventory transfers require the source system to record an 'inbound' transaction, while the downstream financial/inventory system needs to register it as an 'other-outbound (negative)' entry to maintain send-receive balance and ledger-to-physical consistency. This strategy converts source-system transfer-inbound documents into downstream other-outbound documents (with negative quantities) according to defined rules, completing the cross-system inventory business loop.
Data Flow and Field Mapping
Data flow: Source (transfer-inbound business system) → Qeasy Data Integration Platform (middleware transformation layer) → Target (Kingdee Cloud Skyline other-outbound document).
Key field mapping table:
| Business Meaning | Source Field | Middleware Processing | Target Field | Notes |
|---|---|---|---|---|
| Inbound document number | goodsdocNo | Direct pass-through | FJKYNo | Document number |
| Transfer order number | sourceBillNo | Direct pass-through | FJKYLinkNo | Link to upstream document |
| Document type | — | Constant assignment | FBillTypeID | QTCKD95_SYS |
| Inventory organization | goodsDocDetailList_ownerName | Code mapping | FStockOrgId | Centrally maintained |
| Inventory direction | — | Constant assignment | FStockDirect | GENERAL |
| Business date | Source creation time | Format standardization | FDate | yyyy-MM-dd |
The quantity field is negated in the transformation layer, converting positive inbound numbers into negative outbound numbers—this is the core logic of this strategy.
How to Configure on Qeasy
On the Qeasy Data Integration Platform, this strategy is configured as an independent integration flow.
Source configuration: API type is WebAPI (POST), pagination parameters pageIndex/pageSize default to 0/50, incremental pull by creation time window (startDate uses last sync time, endDate uses current time).
Target configuration: Call the batchSave interface to write to Kingdee Cloud Skyline, document type fixed as QTCKD95_SYS, inventory direction fixed as GENERAL.
Centralized code mapping management: Codes for inventory organizations, warehouses, and materials are uniformly maintained in Qeasy's mapping tables to avoid scattering across multiple strategies. A common approach is to first run master data synchronization (customers, suppliers, materials), then run business documents; this strategy should be enabled only after material synchronization is complete.
Implementation Steps
Phased scheduling to avoid overwhelming the target system with full data loads from the start.
Phase 1—Incremental starting point initialization. During first execution, manually specify a historical time point as LAST_SYNC_TIME to ensure incremental pulling starts from that point, avoiding a flood of historical data.
Phase 2—Full trigger (optional). If historical data needs to be backfilled, configure a one-time full-data task in Qeasy to pull inbound documents across the complete time window, transform them, and write to the target. Switch to incremental mode immediately after full-data completion.
Phase 3—Daily scheduling. The source side is recommended to pull every 2 hours (avoiding peak hours), and target writes should be scheduled off-peak (for example, at minute 59 of every hour). Qeasy provides crontab expressions for direct scheduling frequency configuration.
Phase 4—Monitoring and reconciliation. After daily business closes, use Qeasy's run logs to verify whether the source document count matches the target write count; investigate any discrepancies immediately.
Lessons Learned
-
Forgetting to negate quantities: This is the most typical mistake. Transfer-inbound is positive, but in Kingdee other-outbound it must be negative, otherwise inventory ledger-to-physical variance will grow. Be sure to explicitly perform
-qtyprocessing in transformation scripts and add unit tests. -
Code mapping scattered across multiple strategies: If inventory organization and material codes are written separately in each strategy, maintenance costs will be extremely high later. It is recommended to centrally build a mapping table in Qeasy and have all strategies reference the same one.
-
Data loss across scheduling cycle time windows: If source document creation time is near the scheduling boundary, a document may be pulled twice across two cycles or missed entirely. A safe approach is to set startDate 1-2 minutes earlier than the last sync time, and endDate 1-2 minutes later than the current time, leaving a buffer.
-
Dependencies not configured: This strategy depends on master data synchronization such as materials and customers. If upstream master data is not running smoothly when business document synchronization is enabled, target writes will fail or codes will be incorrect.
-
Incorrect full-to-incremental switch timing: After full-data runs complete, if you forget to switch back to incremental, historical data will be repeatedly pulled every day, putting pressure on the target system and generating大量 duplicate documents.
Applicable and Non-Applicable Scenarios
Applicable scenarios: Multi-organization, multi-warehouse retail or manufacturing enterprises that need to synchronize upstream business system transfer-inbound transactions to Kingdee Cloud Skyline for inventory ledger processing. Non-applicable scenarios: Single-organization, single-system enterprises do not need this integration; if the downstream is not Kingdee Cloud Skyline or the document type is not 'other-outbound' mode, field mapping and document type constants need to be redesigned.