Qeasy Cloud
Get Started

Syncing Finished Goods from Kingdee Cloud星辰 to Jushuitan: A Single-Strategy Tutorial

· 吕修远· Integration Solutions· 15 views· 4 min read

What This Strategy Solves

Finished-goods master data is the foundation for ERP and e-commerce fulfillment collaboration. Kingdee Cloud星辰 owns financial and inventory semantics; Jushuitan owns storefront and shipping semantics. Even small drift in SKU codes, specs, or enablement flags causes missing items during order flow and inconsistent inventory views.

This strategy answers one question only: push finished-goods records from Kingdee Cloud星辰 to Jushuitan in a stable, traceable way, serving as the first gate (sequence A) of the broader supply-chain integration.

Data Flow and Field Mapping

Overall direction is B → A: Kingdee Cloud星辰 (source) → Qeasy integration platform (middle layer) → Jushuitan (target).

Business meaningKingdee Cloud星辰 (source)Qeasy middle layerJushuitan (target)
SKU codeItem codesku_code (unified primary key)Item code
SKU nameItem namesku_nameItem name
SpecificationSpecspecSpec
UnitBase unitunitUnit
BarcodeItem barcodebarcodeItem barcode
Active flagDisable flagis_active (normalized boolean)Disabled or not

In real projects we recommend a middle-field layer in Qeasy rather than a rigid source-to-target mapping. This makes future source swaps or new target systems a middle-layer-only change.

How to Configure It in Qeasy

Within the Qeasy data integration platform, this strategy maps to a single strategy node. Key configuration points:

  1. Source connection: Kingdee Cloud星辰 V2 API; credentials are stored in Qeasy's credential vault, not embedded in scripts.
  2. Target connection: Jushuitan open-platform API, also routed through the credential center.
  3. Source data: use the built-in "Kingdee Cloud星辰 finished goods" pull model, supporting incremental pulls by last-modified time and filtering by organization.
  4. Target data: Jushuitan "item profile" write model, defaulting to UPSERT by SKU code.
  5. Field mapping: use the visual mapper to land source fields onto the middle layer, then map from the middle layer to target fields.
  6. Centralized code mapping: enum values like unit and barcode type should live in Qeasy's code-table feature rather than scattered across strategies — one of the most common patterns among Qeasy customers.
  7. Error handling: enable retry and alerting; failed records land in an error log table so manual fixes are auditable.

Implementation Steps

We typically split a finished-goods strategy into three phases to avoid "ran a full sync and crashed with no clue what happened."

  • Phase 1: Incremental baseline. Pick 3–5 representative finished goods in Kingdee (active, disabled, with barcode, without barcode), trigger the Qeasy strategy manually, and verify both ends are identical. This is the baseline.
  • Phase 2: Full trigger. Once baseline passes, run a full sync in Qeasy to push historical finished goods to Jushuitan in one shot. After completion, reconcile total counts, disabled counts, and empty-barcode counts on both sides.
  • Phase 3: Schedule activation. Finished-goods master data does not change often. A typical setup is incremental every 15–30 minutes plus a nightly full sync as a safety net — the "incremental + full dual-track" pattern that many Qeasy customers adopt.

Pitfalls from Real Projects

These are lessons we and our customers have all hit at least once:

  1. Mismatched codes shipped anyway, reconciliation fails months later. A classic mistake is treating Kingdee's code as Jushuitan's code; later when the business needs to rename, both sides change. The safer path is keeping mapping in Qeasy's middle layer so the business only deals with source-system codes.
  2. Active flags not normalized. Kingdee uses 0/1 disable flags; Jushuitan uses true/false. Type mismatches produce dirty data. We normalize to boolean in the middle layer.
  3. Null barcodes treated as errors. Many finished goods have no barcode; the source is null while the target expects an empty string, which can break Jushuitan's non-null validation. Convert null to empty string explicitly in the Qeasy mapper, or skip the barcode field by business rule.
  4. Repeated full syncs triggering Jushuitan rate limits. Jushuitan's API has call-frequency limits; if batch sizes are not tuned, full syncs get throttled. Keep batch sizes and intervals conservative in Qeasy's rate-limit config.
  5. Dependencies not wired. Finished-goods sync must run before downstream strategies like inventory, orders, and pricing. If depends_on is not set in Qeasy, downstream strategies read stale master data. We tag the strategy as sequence A in metadata and enforce scheduling order.

When to Use It — and When Not To

Use it: when multi-org finished-goods master data must be unified into an e-commerce hub, change frequency is low, code consistency is critical, and you plan to add more target systems later. Do not use it: when finished-goods definitions are owned by the e-commerce side and ERP only consumes them, or when field semantics differ so much that bidirectional real-time sync is needed — those cases belong in an MDM initiative, not a single sync strategy.

Original content. Please credit the source when reposting: https://www.qeasy.cloud/insights/solutions/strat-jushuitan-kingdee-cloud-5950-nc7ae6c95-ffb176df

Comments