What data for this category looks like
The core data for this category consists of documents related to ad creative review. Sources include ad delivery platforms, landing page storage services, product qualification filing systems, and public compliance terms. Daily circulating documents include pending review ad creative files, landing page screenshots, rectification opinions issued by reviewers, and iterative records of creative modifications. Data updates occur daily or in real time. Each individual review data entry includes fields such as creative identifier, violation type tag, review timestamp, and associated delivery account. No additional custom statistical fields are included.
Constraints on deployment and upgrades from these characteristics
The real-time or daily update data properties of ad creative reviews require deployments to adapt to high-frequency read-write storage configurations. This prevents review ticket backlogs caused by insufficient storage throughput, which would delay delivery compliance progress. Heterogeneous data from multiple sources (files, text, structured fields) requires deployments to configure parameters compatible with multiple storage types. This ensures normal writing and reading of review data in different formats. The association between review data, delivery accounts, and compliance terms requires that historical review association relationships be retained during upgrades. This avoids loss of historical ticket data or broken associations caused by version upgrades, which creates ad delivery compliance risks. Real-time updated review data also requires upgrade windows to avoid peak business hours. This reduces impact on daily review workflows.
How to set configurations
| Configuration Item | Recommended Value | Rationale |
|---|---|---|
MONGODB_URI | mongodb://username:password@mongodb:27017/fastgpt?authSource=admin | Adapts to the multi-ticket, multi-document storage needs of ad creative reviews, ensuring high-concurrency read-write stability |
REDIS_URL | redis://:redis_password@redis:6379/0 | Used to cache frequently accessed compliance terms and review rules, reducing database query pressure |
DB_MAX_LINK | 200–300 | Matches the concurrent ticket processing requirements of ad creative review scenarios, avoiding request timeouts caused by connection pool exhaustion |
FE_DOMAIN | https://your-fastgpt-domain.com | Used to generate access links and callback addresses for review tickets, ensuring internal operations staff can normally access ticket pages |
STORAGE_S3_ENDPOINT | http://s3-endpoint:9000 | Used to store binary data such as landing page screenshots and ad creative files, adapting to the storage needs of multiple types of review attachments |
Three common mistakes
- Symptom: Post-deployment startup verification fails, with a prompt indicating missing
AIPROXY_API_ENDPOINT(root address of the AI Proxy service; do not append paths such as /v1, not a single-model address) orAIPROXY_API_TOKEN(AI Proxy administrator token; must match the AI Proxy server configuration and cannot be empty, not an application API key or model vendor key). Cause: These two required items are not configured in the environment variables of the App/Admin server, or they are mistakenly filled in workflow global variables, model management pages, or application API request headers instead of service deployment configurations. - Symptom: Image pull fails during Linux server deployment, with a prompt indicating network timeout or image not found. Cause: The correct container image repository address is not configured, or the server does not have permission to access the official image source.
- Symptom: After upgrading to FastGPT v4.17.0, an internal server error prompt appears when importing plugins. Cause: Database indexes or cache configurations were not properly migrated during the upgrade process, causing the plugin to fail to read necessary metadata when loading.
How to confirm the configuration is correct
- Check the environment variables in the App/Admin server deployment configuration. Confirm that
AIPROXY_API_ENDPOINT(root address of the AI Proxy service; do not append paths such as /v1, not a single-model address) andAIPROXY_API_TOKEN(AI Proxy administrator token; must match the AI Proxy server configuration and cannot be empty, not an application API key or model vendor key) are correctly configured and not empty. - Run the service startup verification script to check connectivity to the database, Redis, and AI Proxy.
- Submit a test ad creative review ticket, confirm that the ticket can be normally created, stored, and queried.
- Review service logs, confirm that there are no error messages indicating missing configuration items or connection timeouts.
The values provided are common starting points, and should be evaluated against the relevant deployment 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.