Component Reference

Apache HTTP Server is the world's most widely deployed open-source web server. In the ChainSys architecture it operates exclusively as a reverse proxy and SSL termination point — not as a direct application server.

HTTPD sits at the outermost edge of the DMZ tier, acting as the first layer of ingress control. All inbound traffic from browsers, REST/SOAP clients, and enterprise source systems is received here. HTTPD terminates TLS, enforces HTTPS, and forwards requests into the internal web tier — ensuring no direct internet-to-backend path exists.

HAProxy is a high-performance TCP/HTTP load balancer and proxy, designed for high availability and horizontal scale-out. It is the reference standard for open-source load balancing in enterprise Linux environments.

HAProxy is deployed at every horizontal scale tier in the ChainSys architecture — not just at the web tier. This means the Web Tier (Tomcat cluster), each Application Service cluster (dataZap, dataZen, dataZense, BOTS, SAB independently), and the Database Tier all sit behind a HAProxy instance. New nodes are added to the HAProxy backend pool without downtime.

Keycloak is an open-source Identity and Access Management (IAM) platform supporting SAML 2.0, OAuth 2.0, OpenID Connect (OIDC), LDAP, Kerberos, and Active Directory. It is the ChainSys platform's single authentication authority.

Keycloak is the central IdP for the entire platform. Every authentication flow — user login, service-to-service JWT issuance, SSO, MFA — passes through Keycloak. Zero Trust is enforced at every layer: JWT tokens are validated on every inbound request, with no implicit trust between internal services.

Apache ActiveMQ is a widely deployed open-source message broker supporting JMS, AMQP, MQTT, OpenWire, and STOMP. It provides reliable, asynchronous message delivery with persistence and broker clustering.

ActiveMQ sits in the DMZ as the platform's internal event bus, decoupling producers from consumers across services. It is the backbone of dataZap's CDC (Change Data Capture) pipeline — pipeline trigger events, ETL job notifications, and cross-service coordination messages are delivered through ActiveMQ queues and topics.

Spring Boot is a Java framework that simplifies the bootstrapping and development of Spring-based applications. It provides auto-configuration, embedded server support, and an extensive ecosystem of integrations.

Spring Boot is the current runtime for all ChainSys data and business platform services: dataZap, dataZen, dataZense Catalog, dataZense Analytics, Smart BOTS, Smart App Builder, and the Platform Foundation layer. It has underpinned the platform's growth from inception and remains fully operational during the Quarkus migration.

Quarkus is a Kubernetes-native Java framework built for GraalVM and OpenJDK HotSpot, designed to make Java a first-class citizen in container and cloud-native environments. It features a reactive programming model (Mutiny), compile-time dependency injection (ArC), and native image compilation.

Quarkus is the strategic target runtime for all ChainSys platform services. The AI Gateway is already fully live on Quarkus, validating the stack in production. The migration follows a strangler-fig pattern — existing Spring Boot services remain operational while services are progressively moved to Quarkus.

Apache Tomcat is an open-source Java Servlet container and web server maintained by the Apache Software Foundation. It implements the Jakarta Servlet, Jakarta Expression Language, and WebSocket specifications.

Tomcat forms the Web Tier — the HTTP serving layer that sits between the HTTPD reverse proxy and the application service cluster. It handles HTTP request processing, session management, and routing to the correct application service. The Tomcat cluster is fronted by HAProxy for horizontal scale-out.

Node.js is a JavaScript runtime built on Chrome's V8 engine, designed for building scalable server-side applications. In the ChainSys architecture it is used specifically as the deployment runtime for applications generated by Smart App Builder.

When a developer or citizen developer builds a web or mobile application in the Smart App Builder low-code environment, the generated application is deployed as a Node.js server. The Node.js runtime hosts the app at runtime, connects to CouchDB for app data, and uses dataZap connectors for backend enterprise data access.

PostgreSQL is a powerful open-source object-relational database system with a strong reputation for reliability, feature robustness, and standards compliance. It is the ChainSys platform's primary persistent store for all structured data.

PostgreSQL is used across all five platform services as the source of truth for metadata, workflow state, governance records, datamart results, and platform configuration. It is the most broadly used data component in the platform — every service reads and writes to PostgreSQL.

Apache CouchDB is an open-source document-oriented NoSQL database that uses JSON for documents, JavaScript for MapReduce indexes, and HTTP as its API. Its multi-master replication model makes it well-suited for distributed application data.

CouchDB serves as the dedicated data store for Smart App Builder-generated applications. When SAB generates a web or mobile app, that app's runtime data (user records, form submissions, transactional app data) is stored in a CouchDB partition specific to that app. The reference in the Technical Architecture PDF is explicit: 'CouchDB would be needed only if dynamic mobile applications are to be generated.'

Apache Solr is a highly scalable, open-source enterprise search platform built on Apache Lucene. It provides full-text search, faceted search, hit highlighting, dynamic clustering, and rich document handling.

Solr provides the search and indexing engine for dataZense Catalog exclusively. When data assets are catalogued, their metadata, descriptions, classifications, and PII tags are indexed in Solr. This enables the intelligent search experience in the catalog — keyword search, faceted filtering, and metadata discovery across all catalogued enterprise data assets. The Technical Architecture PDF explicitly states: 'SOLR would be needed only if data cataloging is implemented.'

Redis is an open-source in-memory data structure store, used as a cache, message broker, and session store. It supports strings, hashes, lists, sets, sorted sets, bitmaps, hyperloglogs, and geospatial indexes.

Redis is the platform's shared session cache and API response cache. Because all application services are stateless (a prerequisite for horizontal scaling behind HAProxy), Redis holds the shared state that allows any node in a service cluster to serve any request. This is what makes the ChainSys horizontal scaling model work: nodes are interchangeable because state lives in Redis and PostgreSQL, not in the application process.

Git (distributed) and SVN (centralised) are version control systems for tracking changes to files and coordinating work among developers. In the ChainSys platform they serve as both integration targets (for source data) and storage for platform-managed artifacts.

Git and SVN serve two roles in the platform: as connector targets in dataZap (pulling versioned configuration or code files as source data for migration or integration pipelines), and as the versioning substrate for platform-managed artifacts — ETL mapping templates, transformation scripts, and workflow definitions created within the platform are versioned for auditability and rollback.

The AI Gateway is a ChainSys-built service that brokers all AI/LLM function calls between platform services and external or self-hosted language model providers. It is the first ChainSys service fully built on the Quarkus stack, validating the target architecture in production.

Every AI call within the ChainSys platform — whether from dataZap's AI-assisted data cleansing, dataZen's duplication detection, dataZense's automated metadata generation, or SAB Autonomous agentic workflows — routes through the AI Gateway. No service calls an LLM directly. The Gateway enforces governance, manages System Prompts as versioned artifacts, controls token budgets, and provides a unified audit log of all AI interactions.

Large Language Models (LLMs) are AI systems trained on large text corpora that can perform natural language understanding, generation, reasoning, and code synthesis. ChainSys integrates with both cloud-hosted and self-hosted LLM providers via the AI Gateway.

LLMs power the intelligence layer across all ChainSys products: AI-assisted data cleansing and transformation suggestions in dataZap, duplication detection and golden record scoring in dataZen, automated metadata generation and PII classification in dataZense Catalog, natural language analytics queries in dataZense Analytics, and multi-step agentic workflow orchestration (SAB Autonomous) in Smart App Builder.

A vector store (also called a vector database) stores numerical embeddings — dense vector representations of text, metadata, or other content — and enables similarity search over them. It is foundational for Retrieval-Augmented Generation (RAG) patterns, where LLMs are grounded with retrieved context rather than relying on parametric memory alone.

The vector store is a planned addition to the ChainSys AI architecture to enable semantic search over the Data Catalog and RAG-based intelligent data discovery. Rather than keyword matching (currently served by Apache Solr), embeddings allow finding semantically similar datasets, columns, and business glossary terms — even when exact keywords don't match.

Apache SkyWalking is an open-source Application Performance Monitoring (APM) and observability platform purpose-built for distributed systems. It provides distributed tracing, service topology mapping, metrics aggregation, and log correlation.

SkyWalking provides end-to-end observability across the ChainSys multi-service platform. In a distributed architecture spanning HTTPD, Tomcat, Spring Boot services, Quarkus (AI Gateway), PostgreSQL, Redis, and ActiveMQ, correlating a slow user-facing request back to its root cause requires distributed tracing — SkyWalking provides this across all service boundaries.

Docker is a container runtime for packaging applications and their dependencies into portable container images. Kubernetes (K8s) is the de facto container orchestration platform — managing deployment, scaling, health, and networking of containerised workloads across a cluster.

Kubernetes and Docker are currently deployed for Smart App Builder. The containerisation of SAB validates the deployment model and operational tooling (Helm charts, ingress controllers, autoscaling) before wider rollout. The Quarkus migration is the key enabler for expanding K8s to all other services — Quarkus native images produce containers 60–70% smaller than Spring Boot fat JARs, making per-node density economically viable.