What the Data for This Category Looks Like
Core data for ad creative review includes four categories: ad creative files, landing page screenshots, review opinions, and modification records. Data sources cover ad delivery platforms, landing page hosting services, product qualification filing systems, and advertising law compliance clause libraries.
Data update rhythms fall into two categories: real-time (synced immediately after a single creative is submitted) and daily updates (bulk compliance clause updates).
Single review data includes fields such as creative ID, review status, violation point annotations, reviewer identifier, and modification time. It uses character strings, file sizes, and timestamps as basic identifiers, with no unified fixed unit.
Constraints on Database and Operations
The data characteristics of ad creative review impose multiple constraints on database operations.
Real-time creative and review record updates require databases to support high-concurrency writes and low-latency queries. Reasonable connection pool and index strategies must be configured.
Mixed storage of multiple types of unstructured and structured data requires databases to support both document-based and relational storage solutions, avoiding performance bottlenecks from single storage types.
Associated data from review workflows, such as the binding between creatives and review opinions, requires transaction guarantees to prevent data inconsistency.
Frequently accessed compliance clause and qualification data needs caching to reduce direct database query pressure.
Review errors can lead to risks such as ad rejection or account restriction, requiring databases to have complete backup and disaster recovery mechanisms.
Configuration Settings
These configurations apply to FastGPT v4.17.0.
| Configuration Item | Recommended Value | Rationale |
|---|---|---|
| Database connection pool size | 100–150 connections | Address concurrent query requests for real-time creative review, avoid connection exhaustion |
MONGODB_URI | mongodb://host:port/ads_material?retryWrites=true&w=majority | Store unstructured ad creative metadata and landing page screenshots |
PG_URL | postgresql://user:pass@host:port/audit_records?sslmode=require | Store structured review opinions, modification records, and compliance clause binding data |
DB_MAX_LINK | 600 seconds | Maintain long connections to adapt to real-time creative data transmission |
REDIS_URL | redis://:authpass@host:port/0 | Cache frequently accessed advertising law compliance rules and product qualification data, reduce database load |
| Database backup frequency | 1 daily backup + real-time incremental backup | Ensure disaster recovery and rapid recovery capabilities for review data |
Three Common Mistakes
- Phenomenon: FastGPT displays a connection timeout error when connecting to a locally deployed MongoDB, with
Connection refusedin logs. Cause: The host address and port inMONGODB_URIare not configured correctly, or a local firewall blocks database communication ports. - Phenomenon: Queries for creative reviews return empty results or missing fields. Cause: Corresponding database table models are not created per the business data structure, or field mapping configurations do not match the actual storage structure.
- Phenomenon: Legacy review data cannot load normally after upgrading from v4.6.7 to v4.8.10. Cause: Direct database directory migration was performed without running version compatibility scripts, leading to mismatched database schemas between old and new versions.
How to Verify Proper Configuration
- Run the FastGPT built-in database connection test tool to verify that configuration items including
MONGODB_URIandPG_URLcan establish normal connections. - Import a test ad creative record, confirm that it can be written to the corresponding database table and queried by review status.
- Adjust the
DB_MAX_LINKconfiguration, simulate a multi-concurrency query scenario, and confirm that the number of connections does not exceed the preset threshold and no query timeouts occur. - Trigger a complete creative review workflow, confirm that review records can be synchronously written to the database and corresponding operation logs are generated.
The values provided here are common starting points, and should be evaluated 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.