This page is compatible with FastGPT v4.17.0.
What data for this category looks like
Data for this category primarily originates from backup systems, database operation logs, object storage backup files, and disaster recovery management platforms. Daily produced documents include backup strategy documents, recovery drill reports, disaster recovery switchover plans, and data consistency verification records. Data update frequencies are categorized as daily (backup logs, verification records), monthly (drill reports), and quarterly (switchover plans). Data fields include backup timestamps, recovery duration, verification results, backup file size, and more, with units including seconds, bytes, percentage, and others.
What constraints these characteristics impose on the database and operations link
Multi-source data requires the database and operations link to unify log parsing rules, ensuring backup and recovery data from different systems can be correlated for analysis. Documents with different update frequencies require corresponding batch-wise verification and updates, to avoid executing large-scale drill operations during business peak hours. Diverse document structures require operations configurations to match document formats for different scenarios; for example, recovery drill reports need to correspond to specific database verification fields. The variety of fields requires precise positioning of corresponding fields during fault troubleshooting, to avoid missing key recovery node information.
How to set configurations
| Configuration Item | Recommended Value | Rationale |
|---|---|---|
PG_URL | postgresql://<user>:<pass>@<host>:5432/fastgpt?sslmode=require | Matches Supabase database connection format to ensure access legitimacy |
DB_MAX_LINK | 15–25 | Adapts to multiple concurrent recovery requests during disaster recovery drills, prevents database connection pool exhaustion |
REDIS_URL | redis://:<password>@<host>:6379/0 | Stores temporary verification records and session status for recovery drills, ensures data consistency |
| Recovery Timeout Threshold | 600 seconds | Covers standard disaster recovery recovery durations, avoids false failure judgments for recovery |
| Backup Verification Cycle | 2:00 AM daily | Matches daily data update frequency, does not occupy resources during business peak hours |
MONGODB_URI | mongodb://<user>:<pass>@<host>:27017/fastgpt?authSource=admin | Used to store backup recovery logs and drill reports, complies with FastGPT data storage specifications |
Three common misconfigurations
- Symptom: Returns
text index required for $text queryerror when executing database queries; Cause: No full-text index created for the queried field in the PostgreSQL database, resulting in the post-backup-recovery database being unable to support text retrieval. - Symptom: Knowledge base data lost after FastGPT version upgrade; Cause: No database backup strategy configured in advance, or database mount directory and connection configuration not retained during upgrade.
- Symptom: Data consistency verification failed after disaster recovery drill recovery; Cause: The recovery script did not follow the field order from the backup verification record, resulting in data mapping misalignment.
How to confirm configurations are correctly set
- Access the FastGPT backend database management page, input the configured
PG_URL, and execute a connectivity test to confirm the database connection is normal. - Manually trigger a small-scale recovery drill, review system logs for timeout or connection error records, and verify the rationality of the
DB_MAX_LINKconfiguration. - Check the execution logs of the backup verification task to confirm that the daily verification task completed on schedule without abnormal data.
- Review the cache directory corresponding to
REDIS_URLto confirm that temporary verification data is properly written and cleaned up.
The values provided are common starting points and should be evaluated against operational samples from the relevant deployment.
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.