What this category of data looks like
Core data for this category is maintained daily by operations asset administrators, SREs, and ITSM platform administrators. It includes four types of structured documents: configuration item lists, service component relationship topologies, change records, and asset inventory sheets. Data sources include CMDB systems, cloud resource lists, automated discovery tools, and procurement ledgers. Update frequency is daily or weekly. The most time-consuming steps are configuration item collection, dependency verification, and post-change data updates. This data is directly linked to service fault location and asset account-physical consistency. Unlike the IT asset inventory category, this category focuses on dependencies between service components, and all data is directly traceable structured records.
What constraints do these characteristics impose on the "citation sources and traceability" link
Since data sources are scattered and update rhythms differ, the traceability link must clearly mark the original data source and update time for each cited piece of content, to avoid confusing configuration data from different cycles. The structured nature of configuration items and relationship topologies requires traceability to link specific component IDs and dependency chains, rather than relying solely on text keyword matching. The real-time nature of change records requires traceability results to synchronize the latest configuration modification information, otherwise analysis deviations will occur. At the same time, the risk of asset account-physical inconsistency requires that the traceability link must lock a single original data source, and multiple sets of non-unified configuration data must not be mixed.
How to set the configuration
| Configuration Item | Recommended Value | Rationale |
|---|---|---|
quoteMaxToken | 8000–12000 token | Matches the average content length of CMDB configuration item lists and relationship topologies |
responseData | Return only associated configuration item IDs, change times, and data source tags | Focus on core traceability information and reduce redundant content |
detail | Enabled | Display the complete dependency chain and change history of configuration items |
| Maximum context length | 15000–20000 characters | Cover a single batch of configuration item lists and topology data |
| Single search limit | Top 6 entries | Matches the number of configuration item queries in daily operations |
| Data source filter | Select only CMDB and cloud resource lists | Exclude non-configuration-related procurement ledger data |
This configuration rule applies to FastGPT v4.17.0.
Three common configuration mistakes
- Phenomenon: No citation source for configuration items is displayed in test case analysis results, and the citation field is empty. Cause: The
responseDataparameter is not configured correctly, and the traceability association switch is not enabled. - Phenomenon: API calls return timeout errors, with response duration exceeding
60 seconds. Cause: ThequoteMaxTokensetting exceeds the system processing threshold, resulting in excessively long content parsing time. - Phenomenon: Traceability results include irrelevant procurement ledger entries. Cause: No data source filtering rules are configured, and non-configuration-related document data is mixed.
How to confirm the configuration is complete
- View the traceability pop-up window of the test case, confirm that the associated configuration item IDs, change records, and CMDB data source tags are all displayed correctly.
- Call the FastGPT API interface, check that the content returned in the
responseDatafield complies with the preset filtering rules and contains no redundant data. - Adjust the
quoteMaxTokenvalue, verify that the length of the returned traceability content matches the set parameter range. - Simulate a configuration item change operation, confirm that the traceability result automatically updates to the latest change record.
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.