What the data for this category looks like
Marketing-related data for dedicated equipment comes from hardware such as smart marketing terminals and self-service devices deployed at offline service locations. This data includes device operation logs, marketing material delivery records, and user interaction data.
Two update rhythms apply to this data: Marketing material documents are updated per campaign cycle, typically weekly or monthly. Interaction and operation data reported by devices is updated quasi-real-time, with reports sent every 5 minutes to 1 hour.
Individual data documents contain fields including device_serial, material_code, interaction_count, stay_duration, and report_timestamp. device_serial is a 16-character string. stay_duration uses seconds as its unit. report_timestamp follows the ISO8601 format.
What constraints do these characteristics impose on HTTP interfaces and external systems
Dedicated equipment is typically deployed in private network environments at offline service locations, which have higher latency than public networks. As a result, HTTP interfaces must support longer request timeouts.
The quasi-real-time reporting feature requires interfaces to have high concurrent short connection processing capabilities, to receive reporting requests from multiple devices simultaneously.
Marketing materials include large files such as high-definition images and operational short videos. Interfaces must support large file chunked uploads to reduce transmission failure rates.
Strict field format requirements, such as the 16-character string format for device_serial and the second-level unit for stay_duration, require interfaces to perform field validation on request bodies. This prevents invalid data from entering subsequent processes.
How to set configurations
| Configuration Item | Recommended Setting | Rationale |
|---|---|---|
timeout | 240 seconds | Dedicated equipment is typically deployed in private network environments with higher network latency than public networks. 240 seconds covers most reporting linkages. |
max_body_size | 800–1200 MB | Marketing materials include large files such as high-definition images and operational videos. The maximum single file size can reach 1200 MB, so the upload limit must match this. |
validate_schema | Enabled | Dedicated equipment reporting data must include core fields such as device_serial and material_code. Enabling validation filters invalid reports. |
retry_count | 3 times | Private network environments occasionally experience network fluctuations. 3 retries improve reporting success rates without adding excessive load. |
chunk_mode | Enabled | Uploading large marketing materials is prone to failure due to link interruptions. Chunked uploads reduce the probability of single-transmission failure. |
error_notify_strategy | Configured by location group | Dedicated equipment is distributed across different offline service locations. Configuring by group allows quick localization of the area where faulty devices are located. |
The parameter values provided on this page are common starting points for configuration. Actual values are affected by material formats, data volumes, and business rules. Specific issues require individual analysis. It is recommended to test on your own samples before finalizing settings.
Three common mistakes
- A call to an image analysis interface returns
400 Bad Requestwith a prompt that images cannot be downloaded. This occurs because marketing material images for dedicated equipment are stored in private network private storage, and cross-domain access permissions are not configured. This prevents external interfaces from pulling images. - An HTTP node returns
504 Gateway Timeout, and reported data receives no response. This occurs because the configuredtimeoutvalue is lower than the average latency of the dedicated equipment private network, causing requests to be interrupted early. - Uploaded marketing material files are incomplete, and the
material_contentfield is empty. This occurs becausechunk_modeis not enabled. Link interruptions during large file uploads lead to data truncation.
How to confirm configurations are correct
- Upload a test marketing material file that meets the format requirements, and check that the FastGPT HTTP node returns a
200 OKstatus code. - Simulate a dedicated device reporting test data, and verify that the parsed results of fields such as
device_serialandmaterial_codein the backend logs match the reported values. - Disconnect the network link of the test device, and observe whether the configured group alarm notification is triggered.
- Upload a test file of maximum size, and confirm that the chunked upload process completes normally with no data loss.
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.