What this category's data looks like
Data for this category primarily comes from advertising platforms, asset libraries, user interaction and conversion statistics systems. Daily produced documents include creative scripts, asset test sheets, delivery data reports, and iteration checklists. Data update frequency is primarily daily or weekly. Individual data entries include fields such as creative identifier, asset type, delivery time slot, click-through rate, conversion cost, test cycle, and more. Units include times, yuan, hours, days, and others. Data from different test batches is grouped and aggregated by creative.
Constraints on Database and Operations Workflows
Data sources are distributed across multiple systems. Multiple database connection configurations are required to adapt to different structured and unstructured data formats. The daily or weekly update frequency requires database connections to support stable batch pulling mechanisms. Frequent connection establishment that occupies server resources must be avoided. The requirement to aggregate multi-batch creative test data by creative group requires the database to support multi-dimensional indexing by creative identifier and test cycle. This improves query efficiency across batches. The need for associated analysis of delivery data and asset test data requires ensuring consistency across table associations. Field expansion space must be reserved to adapt to new creative variable dimensions. Data anomalies directly affect delivery effect judgment. Database operations must ensure the stability of data synchronization and the timeliness of queries. Test conclusion deviations caused by data delays or errors must be avoided.
Configuration Recommendations
| Configuration Item | Recommended Value | Rationale |
|---|---|---|
MONGODB_URI | mongodb://username:password@host:27017/fastgpt_ad_creative | Adapt to unstructured test data storage for information flow advertising creative testing, match database instance specifications to business data volume |
PG_URL | postgresql://user:pass@host:5432/ad_creative_db?sslmode=require | Used to store structured asset test sheets and delivery data reports, support multi-dimensional index queries |
DB_MAX_LINK | 20–30 connections | Adapt to concurrent requirements of daily batch data pulling, avoid occupying server resources due to excessive connection count |
REDIS_URL | redis://:password@host:6379/0 | Cache frequently queried creative test metrics, reduce direct database query pressure |
| Database connection timeout period | 30 seconds | Adapt to the response delay range of advertising platform interfaces, avoid data pulling interruptions caused by timeouts |
Three Common Configuration Mistakes
- No output when calling database tools, and the interface returns connection timeout status code 504. The cause is that access permissions for
MONGODB_URIare not configured, preventing FastGPT from establishing a connection with the advertising platform data source. - An error occurs when executing an SQL statement with passed variables, but the same SQL executes without errors manually. The cause is that special characters in the variables are not escaped, damaging the SQL syntax structure.
- FastGPT deployed in a container cannot call database tools. The cause is that the container did not map database ports, and no cross-container network connectivity policy was configured, making it impossible to establish a connection with the database instance.
How to Confirm Configuration Completion
- Log in to the FastGPT v4.17.0 backend database configuration page, verify that the connection status of
MONGODB_URIandPG_URLshows as normal. - Create a new test workflow, pass variables containing special characters to execute an SQL query, confirm that the returned result matches the manual execution result.
- View the FastGPT container runtime logs, confirm there are no error messages about database connection failures or timeouts.
- Simulate a daily data pulling task, confirm that the database can normally pull advertising platform test data in batches.
The values provided are common starting points and should be measured against the reader's own samples.
Question material comes from public community discussions. Configuration values are common starting points and should be measured against your own samples. Verified on 2026-10-19.