What the data for this category looks like
Duty-free financing daily report data is sourced from daily operating cash flows, supplier credit ledgers, and settlement systems of offline outlying-island duty-free shops and urban duty-free shops. Full synchronization of the previous day’s data runs daily between 1:00 AM and 3:00 AM. Each daily report document uses store codes as the top-level grouping identifier. Each group contains five core fields:
- Daily duty-free merchandise sales revenue
- Daily available credit limit
- Daily financing received amount
- Accounts payable to suppliers
- Settlement period
Amount fields use Chinese Yuan (CNY) as the unit. Settlement period uses calendar days as the unit. There are no nested sub-document levels. All fields are required and cannot be left empty.
Constraints Imposed on HTTP Interfaces and External Systems by These Characteristics
The fixed daily update schedule requires HTTP interfaces to support scheduled pull scheduling configurations, and to handle full daily batch data returns to avoid omissions from incremental pulls. All fields are required and cannot be empty. Interface request parameter validation must strictly match field names and data types to prevent synchronization failures caused by missing fields. External systems must complete unit conversion adaptation before integration, to align with the fixed units for amount and settlement period fields, avoiding mismatched currency or settlement period units. The non-nested document structure requires interface return data to use a flat structure. Interfaces must also support pagination pulls by store code to accommodate data pull requirements for duty-free shops of varying sizes.
How to Configure Settings
| Configuration Item | Recommended Value | Rationale |
|---|---|---|
API_REQUEST_TIMEOUT | 600 seconds | Adapts to the pull duration for full daily report data, prevents request interruptions caused by large data volumes |
MAX_RESPONSE_BATCH_SIZE | 1000 items | Matches the maximum number of stores per daily report, prevents single return data from exceeding system processing thresholds |
FIELD_VALIDATION_STRICTNESS | Strict validation | Aligns with the required non-empty, mandatory field characteristics of duty-free financing daily reports, filters invalid data |
CERT_MOUNT_PATH | /etc/fastGPT/certs | Adapts to the standard certificate mount path for HTTPS deployments, meets general deployment requirements |
API_PAGINATION_PAGE_SIZE | 50 items | Balances interface response speed and data pull efficiency, accommodates integration scenarios for duty-free shops of varying sizes |
The parameter values provided on this page are common recommended starting points for configuration. Actual values are affected by material form, data volume, and business rules. Each situation requires specific analysis. It is recommended to test on your own samples before finalizing values.
Three Common Mistakes
- Symptom: After starting the container, the interface cannot be accessed via HTTPS, returning a
400 Bad Requesterror. Cause: The certificate mount path is not correctly configured in docker-compose.yml, causing FastGPT to fail to read the SSL certificate files. - Symptom: Store grouping fields are missing or directory levels display abnormally in API-returned daily report data. Cause: The interface return data structure is not flattened, with nested levels that do not match FastGPT’s parsing rules.
- Symptom: Knowledge base content can be retrieved normally on the debug page, but some questions fail to match the knowledge base via API calls. Cause: The
conversationIdorsystemPromptparameters are not correctly passed during API calls, causing context to differ from the debug environment.
How to Confirm Proper Configuration
- Initiate a full data pull request. Verify that returned fields include the preset core fields, and that field types and units match duty-free financing daily report requirements.
- Review container runtime logs to confirm the HTTPS certificate has loaded successfully, with no
certificate not founderror messages. - Call the API interface with a test question. Compare retrieval results from the debug page and the API to confirm retrieval logic matches the debug environment.
- Simulate a scheduled pull task. Confirm data synchronization triggers automatically at the preset time, with no delays or interruptions.
Question material comes from public community discussions. Configuration values are common starting points and should be measured against your own samples. Verified on 2026-09-14.