IRE — Identity &
Reconciliation Engine
Discovery sends data to the CMDB — IRE decides what to do with it. Every CI write goes through IRE's three-phase pipeline: Identification (does this CI already exist?), Reconciliation (who owns each field?), and Deduplication (are there duplicate CIs?). IRE is the engine that maintains CMDB integrity in a multi-source environment.
Critical exam traps: IRE API vs. Table API (always use IRE API for integrations — Table API bypasses IRE), Identification Rule entry logic (AND logic, missing field = entry skipped), Reconciliation Rule staleness period (per field, not per CI), and Deduplication (fix root cause before running remediation playbooks — don't merge duplicates before fixing the Identification Rule). The IRE Log (cmdb_ire_event) is the primary diagnostic tool for all ingestion issues.
Part 4 Study Modules
IRE Architecture — The Three-Phase Pipeline
IRE's role as CMDB gatekeeper, three-phase pipeline (Identification → Reconciliation → Deduplication), IRE API vs. Table API (critical exam trap), source labels, and IRE Log (cmdb_ire_event) as primary diagnostic tool.
Identification Rules — Does This CI Already Exist?
Identification Rule anatomy, Identifier Entries (AND logic — all fields must match), out-of-box identification rules by CI class (serial_number+manufacturer+model_id for servers), Lookup Rules for dependent CIs, Test Identification utility.
Reconciliation Rules — Who Owns Each Field?
Reconciliation Rule anatomy, CMDB Data Source Priority (Manual > Discovery > SGC > Other), field ownership modes (Fully Authoritative, Priority-Based, First-in Wins), Staleness Period (per field — not per CI), discovery_source field, null value handling.
Deduplication & Remediation — Resolving Duplicate CIs
How duplicates form (4 common causes), IRE Phase 3 real-time dedup, cmdb_dedup table, Master CI determination criteria, Remediation Playbooks (Merge, Keep Primary, Flag, Dismiss), merge process flow, and the critical rule: fix root cause before running remediation.