Agentic AI Reference Architecture

Infrastructure · Software component

Shard Replication Manager

Software componentInfrastructureInfrastructurearc:ShardReplicationManager

A cluster component that keeps each data shard on the configured number of nodes, continues writes to surviving replicas during failures and resynchronizes nodes when they rejoin.

Responsibility. Maintains shard replicas and catch-up resynchronization.

deployed onwritesis triggered byis configured byContainer Orchestrator: deployed onContainer OrchestratorSelf-Managed Vector Index Store: writesSelf-Managed Vector Inde…Gossip Membership Service: is triggered byGossip Membership ServiceReplication and Sharding Configuration: is configured byReplication and Sharding…
Direct neighbourhood (hover for relationship types)

Relationships

deployed on structural

is configured by structural

writes dependency

is triggered by dynamic

Classification

Patterns
Replication factor NPending-resync marking during node outageCatch-up replication on rejoin
Quality attributes
Reliability (ISO/IEC 25010 | NIST AI RMF: valid and reliable)
Risks mitigated
Single-node outagesManual intervention during brief failures

Sources

  1. Ch6.2B: T. Nguyen, "Production Vector Database Deployment," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 6.2B. ISBN: 9798244538229.