MongoDB Enterprise Java Examples

This monorepo contains three sibling projects that together demonstrate enterprise-grade MongoDB patterns in Java:

memex - MongoDB Enterprise Microservice Examples

A Spring Boot 3.4 / Java 17 microservice that showcases production MongoDB patterns: streaming bulk-load with upsert strategies (REPLACE, UPDATE, UPDATEWITHHISTORY, INSERT), ACID-transaction change-history, Atlas Search, Kafka ingestion, dynamic config-driven query endpoints with query-costing, and downstream streaming with native BSON-to-JSON conversion.

Build: mvn clean package in memex/ -> memex-1.0.jar

DataGen - Probabilistic Data Generator

A standalone Java 17 CLI that generates large volumes of realistic JSONL test data from a directory of weighted CSV (or .csv.gz) definition files. Supports auto-incrementing IDs (@ONEUP), random integers/doubles/dates (@INTEGER, @DOUBLE, @DATE, @DATETIME), nested arrays (@ARRAY(subdir,n)), embedded literal JSON, and parallel sharded generation. The output is streamable JSONL ready for mongoimport or memex's bulk-load endpoints.

java -jar DataGen-1.0.jar <inputDir> <docsToGenerate> <outputFile>

mxtest - Minimal HTTP API Test Client

A single-file Java 17 CLI (267 lines) for manually exercising REST endpoints. It sends GET/POST/DELETE requests, pretty-prints JSON responses, supports field-filtering via dotted paths, and loads POST bodies from files. Its only dependency is Gson. Think of it as a lightweight, JSON-aware curl for quick ad-hoc testing of memex endpoints.

java -jar mxtest-1.0.jar GET|POST|DELETE <URL> [field1,field2...]

How they fit together

DataGen --generates--> JSONL data --POST /api/inspections--> memex --MongoDB-->
                                                                ^
                                                             tested via
                                                                |
                                                            mxtest (HTTP CLI)

• DataGen creates statistically realistic JSON datasets
• memex runs as a Spring Boot service, loading that data into MongoDB and demonstrating enterprise patterns (streaming, history, search, Kafka, etc.)
• mxtest is a lightweight HTTP client for manually hitting memex's REST endpoints and inspecting responses

Workshop materials

The original SAMPLE_DATA/ (UK VOSA MOT dataset scripts), Notes/, and sample JSON files have been moved to the MEMEX_Workshop private repo. companies250k.json (403MB raw) is stored there as companies250k.json.gz to stay under GitHub's file size limit.