@biorate/migrations

@biorate/migrations

Database and infrastructure migration framework — manages schema and data migrations across Sequelize, MongoDB, Minio, Kafka, Clickhouse, AMQP, and Schema Registry.

Features

• Multi-engine — RDBMS (Sequelize), MongoDB, Minio/S3, Kafka Admin, Clickhouse, AMQP/RabbitMQ, Schema Registry.
• Tracking — tracks completed migrations per engine (Sequelize/MongoDB/Clickhouse store in a table/collection).
• JavaScript/SQL — JS scripts for complex logic, SQL files for schema DDL.
• DI-orchestrated — all connectors managed via @biorate/inversion container.
• Side-effect import — @biorate/migrations auto-runs when loaded (node -r @biorate/migrations).

Installation

pnpm add @biorate/migrations

Requires the underlying engine packages (@biorate/sequelize, @biorate/mongodb, @biorate/minio, @biorate/kafkajs, @biorate/clickhouse, @biorate/amqp, @biorate/schema-registry, @biorate/vault).

Quick start

// package.json
{
  "scripts": {
    "migrations": "node -r @biorate/migrations"
  }
}

Run with environment config:

MIGRATIONS_CONFIG=./config.ts npm run migrations

Module reference

Root — Orchestrator

Root is the main DI class that initialises all migration connectors. Binds every connector into the @biorate/inversion container and calls process.exit(0) on completion.

import { Root } from '@biorate/migrations';

// Root is auto-instantiated on package import.
// Manually:
import { container } from '@biorate/inversion';
const root = container.get<Root>(Root);
await root.$run();

Migration — Abstract base

import { Migration } from '@biorate/migrations';

Abstract base class for all migration types. Provides:

Member	Signature	Description
config	protected config: IConfig	Injected config service.
connector	protected abstract connector: IConnector	Engine connector (must be overridden).
type	protected get type(): string	this.constructor.name.toLowerCase().
scan(...args)	async scan(...paths): Promise<string[]>	Scans migration directory, returns file paths.
path(...args)	path(...segments): string	Builds path: cwd/migrations/{type}/{...}.
log(...args)	log(...msg): void	Console logging with type prefix.
forEach(namespace, callback)	async forEach<T, C>(ns, fn)	Iterates config entries and connections.
forEachPath(paths, callback)	async forEachPath(paths, fn)	Iterates files in a path list.
process()	protected abstract process(): Promise<void>	Override with migration logic.
initialize()	@init() protected async initialize()	Calls this.process() on lifecycle init.

Sequelize — RDBMS migrations

import { Sequelize } from '@biorate/migrations';

Reads SQL files (e.g. 00001_create_users.sql) from migrations/sequelize/, executes them in transactions, and tracks completion in a migrations table (name CHAR PRIMARY KEY).

Config namespace: Sequelize (from @biorate/sequelize base config).

Directory: ./migrations/sequelize/

Mongodb — MongoDB migrations

import { Mongodb } from '@biorate/migrations';

Loads and executes JS scripts (module.exports = async (connection, config, globalConfig) => {...}) from migrations/mongodb/. Tracks completed scripts via documents in a migrations collection.

Config namespace: MongoDB (from @biorate/mongodb base config).

Directory: ./migrations/mongodb/

Minio — Minio/S3 migrations

import { Minio } from '@biorate/migrations';

Loads and executes JS scripts (module.exports = async (connection, config, globalConfig) => {...}) from migrations/minio/. Tracks completed scripts via marker objects in a migrations bucket.

Config namespace: Minio (from @biorate/minio base config).

Directory: ./migrations/minio/

Kafka — Kafka Admin migrations

import { Kafka } from '@biorate/migrations';

Loads and executes JS scripts (module.exports = async (connection, config, globalConfig) => {...}) from migrations/kafka/. Creates/alters topics, configs, etc. No tracking (every run re-applies).

Config namespace: KafkaJSAdmin (from @biorate/kafkajs KafkaJSAdminConnector).

Directory: ./migrations/kafka/

Clickhouse — Clickhouse migrations

import { Clickhouse } from '@biorate/migrations';

Reads SQL files from migrations/clickhouse/, executes them via connection.command(), and tracks completion in a migrations MergeTree table with name String primary key.

Config namespace: Clickhouse (from @biorate/clickhouse base config).

Directory: ./migrations/clickhouse/

Amqp — AMQP/RabbitMQ migrations

import { Amqp } from '@biorate/migrations';

Loads and executes JS scripts (module.exports = async (channel, connection, config, globalConfig) => {...}) from migrations/amqp/. The script receives the AMQP channel for declaring exchanges/queues/bindings. No tracking.

Config namespace: Amqp (from @biorate/amqp base config). Additional config: migrations.Amqp.<name>.amqpChannelOptions.

Directory: ./migrations/amqp/

SchemaRegistry — Schema Registry migrations

import { SchemaRegistry } from '@biorate/migrations';

Reads .avsc.json / .json files named 00001_subjectname.avsc.json from migrations/schema-registry/. Each file is registered as a new subject or a new version of an existing subject on the Schema Registry.

Feature	Description
File pattern	00001_{subject}.avsc.json
Compatibility	Set via config, defaults to FORWARD
On conflict	Registers a new version for the subject
Errors	SchemaRegistryWrongFileNameError if pattern mismatch

Config namespace: SchemaRegistry (from @biorate/schema-registry base config).

Directory: ./migrations/schema-registry/

Errors

Error	Context
SchemaRegistryWrongFileNameError	Schema Registry migration file name does not match 00001_name.avsc.json pattern.

Config structure

Your config file (referenced by MIGRATIONS_CONFIG env var) must define the connection configs for all engines you use, for example:

export default {
  Sequelize: [{ name: 'primary', options: { dialect: 'postgres', ... } }],
  MongoDB: [{ name: 'primary', host: 'mongodb://...', options: { dbName: 'test' } }],
  Minio: [{ name: 'primary', options: { endPoint: 'localhost', ... } }],
  KafkaJSAdmin: [{ name: 'primary', options: { 'metadata.broker.list': 'localhost:9092' } }],
  Clickhouse: [{ name: 'primary', options: { host: 'localhost', ... } }],
  Amqp: [{ name: 'primary', options: { hostname: 'localhost', ... } }],
  SchemaRegistry: [{ name: 'primary', options: { baseURL: 'http://localhost:8081' } }],
};

Architecture

@biorate/migrations

MIGRATIONS_CONFIG / MIGRATIONS_ROOT env → requireCjs
│
└── Root (DI orchestrator)
    ├── container.bind(SequelizeConnector, MinioConnector, MongoDBConnector, ...)
    │
    └── @init() initialize()
        ├── Sequelize.process()
        │   ├── forEach config → connection.query(sqlFile)
        │   └── INSERT INTO migrations (name) VALUES (...)
        │
        ├── Mongodb.process()
        │   ├── forEach config → require(jsFile)(connection, config, globalConfig)
        │   └── db.migrations.insertOne({ _id: name })
        │
        ├── Minio.process()
        │   ├── forEach config → require(jsFile)(connection, config, globalConfig)
        │   └── client.putObject(migrationBucket, name, ...)
        │
        ├── Kafka.process()
        │   └── forEach config → require(jsFile)(adminClient, config, globalConfig)
        │
        ├── Clickhouse.process()
        │   ├── forEach config → connection.command(sqlFile)
        │   └── INSERT INTO migrations (name) VALUES (...)
        │
        ├── Amqp.process()
        │   └── forEach config → require(jsFile)(channel, connection, config, globalConfig)
        │
        └── SchemaRegistry.process()
            └── forEach config → POST /subjects/{subject}/versions
                └── SchemaRegistryWrongFileNameError check

Learn

• Documentation can be found here - docs.

Release History

See the CHANGELOG

License

MIT

Copyright (c) 2021-present Leonid Levkin (llevkin)