Qeasy Cloud
Get Started

Warehouse Master Data Sync in Practice: UFIDA NCC → Jikeyun — Single-Strategy Deep Dive

· 卢剑航· Integration Solutions· 8 views· 4 min read
吉客云用友NCC仓库主数据供应链集成轻易云Incremental Sync

What This Strategy Solves

In a project spanning both retail and manufacturing, the customer runs UFIDA NCC as the back-end for finance and supply chain, while stores and e-commerce fulfillment use Jikeyun. In the first week, the warehouse supervisor noticed that warehouse records were misaligned across the two systems: the same warehouse was "Beijing Warehouse 1" in UFIDA, but "BJ01" in Jikeyun. With manual mapping tables maintained by people, the gap only widened over three months.

The goal of this strategy is straightforward — treat UFIDA NCC as the single source of truth for warehouse master data and push it into Jikeyun so downstream inventory documents can route to the correct warehouse automatically. Our approach is to use the Qeasy data integration platform to host the entire pipeline, with all code mappings managed centrally in the middle layer, avoiding drift caused by each side maintaining its own rules.

Data Flow and Field Mapping

The pipeline is a one-way push: UFIDA NCC (source) → Qeasy middle layer → Jikeyun (target).

Business MeaningUFIDA NCC FieldQeasy Middle LayerJikeyun Field
Warehouse codewarehouse_codewh_code (standard code)warehouse_code
Warehouse namewarehouse_namewh_namewarehouse_name
Warehouse typewh_typewh_type_norm (normalized enum)type
Active statusis_activestatus (0/1)enabled
Owning organizationorg_codeorg_reforg_code
Last modified timemodify_datelast_modifiedupdate_time

The most critical action in the middle layer is centralized code mapping management: convergence of UFIDA organization codes, warehouse-type enums, and similar values into mapping tables on Qeasy, rather than scattering the logic across per-strategy scripts. This way, subsequent strategies of the same kind can reuse the same mapping table, and a change in one place takes effect everywhere.

How to Configure on Qeasy

Open the Qeasy integration platform and locate the integration solution that contains this strategy, then enter the strategy editor.

  1. Source connector: Select the UFIDA NCC OpenAPI connector. Configure access credentials for the accounting book and limit the interface scope (only tick the warehouse-archive-related endpoints, following the least-privilege principle).
  2. Target connector: Select Jikeyun's warehouse write interface and confirm it is incremental write rather than overwrite.
  3. Data source configuration: Use Qeasy's source API fetch view, with the filter condition set to modify_date >= ${LAST_RUN_TIME} to prepare for incremental runs.
  4. Field mapping: In the mapping canvas, drag and drop the fields listed in the table above. Enum values are handled by a "code conversion" component that references entries in the mapping table.
  5. Write policy: Tick "decide based on primary key existence" — update if the record exists, insert if not — to avoid duplicates.
  6. Exception handling: Enable Qeasy's failure retry and alert channels, and write errors to the unified log for troubleshooting.

The whole configuration does not require complex scripts. Qeasy's drag-and-drop canvas combined with componentized field conversion covers around 80% of warehouse master data sync scenarios.

Implementation Steps

For warehouse master data synchronization, we typically recommend a three-step approach:

Step 1: Full initialization. Before project go-live, pull all warehouse records from UFIDA NCC in one pass and push them into Jikeyun to align both sides. This is typically done with Qeasy's "full trigger" action. Once complete, immediately switch the strategy to incremental mode.

Step 2: Align the incremental starting point. Before switching to incremental, make sure to anchor Qeasy's LAST_RUN_TIME to the moment the full run completed, to avoid data loss in the window in between. This is a classic pitfall.

Step 3: Daily scheduling. Warehouse master data does not change frequently. We generally set the schedule to every 30 minutes, which is real-time enough without putting pressure on either system. If factory warehouses and store warehouses are managed separately, you can also schedule them in parallel by warehouse type.

In addition, on Qeasy we enable a dual-track mode of incremental and full: incremental runs day to day, and once a month a full validation pass runs to detect any dirty data that may have accumulated during incremental runs.

Lessons Learned

1. Code mappings scattered across scripts. The first time we did this, we wrote organization code conversion into each strategy's script, resulting in three different sets of logic across three projects. The safe approach is to extract all enum mappings into Qeasy's unified "code conversion" management and only reference them from strategies, never redefine.

2. Forgetting to switch from full to incremental after the full run. Once the full run completes, the strategy still defaults to running as a full run by default, treating each schedule as a complete rewrite. Jikeyun's interface then sees sudden pressure. Always mark a "full run completed" flag explicitly on Qeasy, and let subsequent schedules go incremental automatically.

3. The active status flag is overlooked. UFIDA's "disabled warehouse" has no corresponding field in Jikeyun. If you do not handle this, inventory in disabled warehouses will still flow. A common mistake is to write the status across directly, but the target field is empty, causing inconsistent status on both sides. The safe approach is to skip the push or mark it as disabled in the middle layer when is_active=0.

4. Field type mismatch. UFIDA's modify_date is a string, while Jikeyun's update_time requires a timestamp. A direct push will fail on format. Make sure to add a time-formatting step in the mapping component.

5. No idempotency check. Warehouse master data does not change at high frequency, but occasional reruns may produce duplicate requests. Enable Qeasy's "decide based on primary key existence" write policy to avoid repeated insertions of the same warehouse.

Applicable and Inapplicable Scenarios

Applicable: Multi-organization, multi-warehouse-type scenarios where ERP must serve as the single source of truth for warehouse archives and push them to downstream WMS or e-commerce fulfillment systems; warehouse change frequency is moderate (from a few to a few dozen times per day).

Not applicable: Scenarios where warehouse archives are co-maintained across multiple systems and need to be edited on both sides — one-way sync will overwrite manual edits on the target side. Also not applicable for inventory scheduling with extremely high real-time requirements (second-level), since master data sync is typically used as a prerequisite foundation.

Original content. Please credit the source when reposting: https://www.qeasy.cloud/insights/solutions/strat-p2ea595-ncc-3109-ok-1c4c2724

Comments