What the data for this category looks like
The daily revenue and market report data for the thermal category is sourced from the real-time collection system of the thermal operation platform and the enterprise financial settlement module. It updates once daily, with aggregation and generation of the previous day’s data completed each early morning. Each daily report document is a structured record containing these core fields: total daily heating scale (unit: gigajoule), unit heating cost (unit: yuan/gigajoule), average daily heating duration (unit: hours), number of terminal user payment transactions that day, and number of operation and maintenance anomalies that day. There are no redundant fields, only core indicators covering daily operations and revenue. The data uses a unified JSON or CSV format to facilitate batch import and parsing.
Constraints imposed on deployment and upgrade
The characteristics of the daily report data for the thermal category impose three core constraints on deployment and upgrade. First, the fixed daily update cadence requires configuring a scheduled task trigger rule that matches the data source synchronization rhythm, to avoid retrieving delayed or expired data. Second, the multi-dimensional mixed fields (related to operation scale, cost, and revenue) require clear vector database field mapping rules during deployment, to ensure different dimension indicators can be associated during retrieval. Third, the incremental per-record data scale requires configuring appropriate batch import parameters during deployment, to avoid service timeouts caused by overly large single import data volumes. The upgrade process must retain the incremental synchronization configuration logic, to avoid daily report tasks being interrupted by full re-imports. It must also support gray release to ensure service continuity.
How to configure the settings
| Configuration Item | Recommended Value | Rationale |
|---|---|---|
SCHEDULE_CRON | 0 2 * * * | Matches the update completion time of most thermal data sources for the previous day, ensuring access to the latest daily report data |
VECTOR_FIELD_MAPPING | {"total_heating_scale": "heat_total", "unit_cost": "unit_cost", "payment_count": "pay_count"} | Corresponds to the core fields of the daily report document, ensuring that vector retrieval can associate operational and financial dimension data |
BATCH_IMPORT_SIZE | 500 entries per call | Adapts to the incremental synchronization scale of daily single-record reports, avoiding timeouts caused by overly large single import data volumes |
API_KEY_REQUIRED | true | Restricts access to revenue and market reporting results only to authorized users, complying with the sensitivity requirements of thermal operation data |
SANDBOX_IMAGE_PULL_POLICY | IfNotPresent | Ensures the upgrade process prioritizes locally cached images, reduces pull time, and avoids upgrade interruptions |
PARSE_DATA_TIMEOUT | 300 seconds | Adapts to the field parsing duration of daily report documents, avoiding parsing timeouts caused by data volume fluctuations |
The parameter values provided on this page are common starting points for configuration. Actual values are affected by material form, data volume and business rules. Specific issues require case-by-case analysis. It is recommended to test on your own samples before finalizing settings.
Common configuration mistakes
- Issue: After configuring
API_KEY_REQUIREDastrue, calling the reporting interface returns a 401 Unauthorized error. Cause: The correctX-API-Keyparameter is not included in the request header, or the parameter value does not match the authentication key configured during deployment. - Issue: After upgrading the version, executing the daily report generation workflow fails to retrieve thermal revenue data output by the code run. Cause: The workflow’s variable mapping configuration was not retained during the upgrade process, causing the output fields of the code node to not be correctly associated with subsequent reporting nodes.
- Issue: Pulling the
fastgpt-sandboximage fails during Docker deployment, with the console displaying a message that the image does not exist or the pull timed out. Cause: The correct image repository address was not configured, or image pull permissions were not enabled, preventing retrieval of the officially maintained sandbox image.
How to verify successful configuration
- Manually trigger a scheduled task, check that the generated daily report document includes all configured fields and that the data matches the actual records of that day’s thermal operations.
- Call the reporting interface with correct authentication parameters, confirm that no 401 error is returned, and that the returned fields are complete and their values match the configured mapping rules.
- View the Docker container’s running logs, confirm that the
fastgpt-sandboximage was successfully pulled and started, with no logs indicating pull failure or startup abnormalities. - Connect to the vector database, retrieve the configured core fields, confirm that the index has correctly mapped all configured items and can return associated multi-dimensional data.
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.