arcxa

 

In ARCXA, data governance, schema mapping, and data migrations are managed via an RDF-style Triple Store Architecture handled by its dedicated data plane component, arcxa-shard.

Instead of traditional relational tables (rows and columns), ARCXA breaks down every single piece of information, relationship, policy, and data lineage trace into granular, atomic statements called Triples.

[Subject - Predicate - Object] process works under the hood in ARCXA to model and move data.

1. The Core Anatomy: What is a Triple?

A triple is a semantic statement that mimics natural language structure. By chaining millions of these statements together, ARCXA builds a massive, highly queryable Knowledge Graph.

• Subject: The entity or resource being described (e.g., a database column, a system interface, or a data entity). In ARCXA, this is uniquely identified using a URI.

• Predicate: The specific relationship, property, or action connecting the Subject to the Object. It defines how they interact.

• Object: The value or the secondary entity being pointed to. This can be another unique resource (a node) or a raw data value/literal (like a string or a number).

A Practical ARCXA Lineage Example:

Imagine tracking where a piece of customer data came from during an enterprise migration. ARCXA models it like this:

Subject (Entity)	Predicate (Relationship)	Object (Value/Entity)
arcxa://dataset/legacy_crm	has_column	arcxa://column/cust_id
arcxa://column/cust_id	maps_to_ontology	arcxa://ontology/CustomerID
arcxa://workflow/nightly_sync	transformed	arcxa://dataset/legacy_crm

2. How the Process Works inside ARCXA

ArcXA architecture processes these triples through a strict workflow handled between the control plane (arcxa-coordinator) and the data plane (arcxa-shard).

Step 1: Ingestion and Semantic Mapping

When you connect an operational data source to ARCXA, the framework parses the metadata. Instead of just copying the schema, it uses its semantic mapping engine to translate the database’s structural rules into triples.

• For example: A database foreign key between Orders and Customers is automatically ingested and stored as a triple: [Orders] -> [belongs_to] -> [Customers].

Step 2: Storage and Distributed Indexing via Shards

Once the triples are generated, they are fed into arcxa-shard, ARCXA's specialized distributed RDF data plane.

• To make querying lightning-fast, ARCXA indexes these triples in multiple combinations—typically combinations of SPO (Subject-Predicate-Object), POS (Predicate-Object-Subject), and OSP (Object-Subject-Predicate).

• This allows the database to instantly answer questions from any angle (e.g., "Find all workflows that touched Dataset X" vs. "Find all datasets touched by Workflow Y").