Skip to content

Latest commit

 

History

History
241 lines (176 loc) · 5.68 KB

README.md

File metadata and controls

241 lines (176 loc) · 5.68 KB

Build Status NPM Status Codecov Status

Momy

Momy is a simple cli tool for replicating MongoDB to MySQL in realtime.

  • Enable SQL query on data in NoSQL database
  • Enable to be accessed by Excel / Access

Momy

Installation

Install via npm:

$ npm install -g momy

Or use docker:

$ docker run -it --rm -v $(pwd):/workdir cognitom/momy

You might want to create an alias, for example

$ echo 'alias momy="docker run -it --rm -v $(pwd):/workdir cognitom/momy"' >> ~/.bashrc

See more detail about Docker configurations below.

Preparation

MongoDB

Momy uses Replica Set feature in MongoDB. But you don't have to replicate between MongoDB actually. Just follow the steps below.

Start a new mongo instance with no data:

$ mongod --replSet "rs0" --oplogSize 100

Open another terminal, and go to MongoDB Shell:

$ mongo
....
> rs.initiate()

rs.initiate() command prepare the collections that is needed for replication.

MySQL

Launch MySQL instance, and create the new database to use. The tables will be created or updated when syncing. You'll see mongo_to_mysql, too. This is needed to store the information for syncing. (don't remove it)

Configuration

Create a new momyfile.json file like this:

{
  "src": "mongodb://localhost:27017/dbname",
  "dist": "mysql://root:password@localhost:3306/dbname",
  "prefix": "t_",
  "case": "camel",
  "collections": {
    "collection1": {
      "_id": "number",
      "createdAt": "DATETIME",
      "field1": "number",
      "field2": "string",
      "field3": "boolean",
      "field4.subfield": "string"
    },
    "collection2": {
      "_id": "string",
      "createdAt": "DATETIME",
      "field1": "number",
      "field2": "string",
      "field3": "boolean",
      "field4": "TEXT"
    }
  }
}
  • src: the URL of the MongoDB server
  • dist: the URL of the MySQL server
  • prefix: optional prefix for table name. The name of the table would be t_collection1 in the example above.
  • fieldCase: optional. snake or camel. See the section below.
  • exclusions: optional. Chars or a range of chars to exclude: "\uFFFD"
  • inclusions: optional. Chars or a range of chars to include: "\u0000-\u007F"
  • collections: set the collections and fields to sync

_id field is required for each collection and should be string or number.

Field names and types

"<field_name>": "<field_tipe>"

or, field_name could be dot-concatenated:

"<field_name>.<sub_name>": "<field_tipe>"

For example, if you have { a: { b: { c: 'hey!' } } } then "a.b.c": "string"

Currently these native types are supported:

  • BIGINT
  • TINYINT
  • VARCHAR
  • DATE
  • DATETIME
  • TIME
  • TEXT

There're also some aliases:

  • number => BIGINT
  • boolean => TINYINT
  • string => VARCHAR

Field name normalization: fieldCase

Some system like Microsoft Access don't allow dot-concatenated field names, so address.street will cause an error. For such a case, use fieldCase:

  • snake: address.street --> address_street
  • camel: address.street --> addressStreet

Note: if you set fieldCase value, the name of _id field will change into id without _, too.

Usage

At the first run, we need to import all the data from MongoDB:

$ momy --config momyfile.json --import

Then start the daemon to streaming data:

$ momy --config momyfile.json

or

$ forever momy --config momyfile.json

Usage with Docker

First thing first, create a network for your containers:

$ docker network create my-net

Then, launch database servers:

$ docker run \
    --name my-mongod \
    --detach --rm \
    --network my-net \
    --mount type=volume,source=my-mongo-store,target=/data/db \
    mongo --replSet "rs0"
$ docker run \
    --name my-mysqld \
    --detach --rm \
    --network my-net \
    --mount type=volume,source=my-mysql-store,target=/var/lib/mysql \
    --env MYSQL_ALLOW_EMPTY_PASSWORD=yes \
    mysql

If this is the first time to run the containers above, you need to initialize them:

$ docker exec my-mongod mongo --eval 'rs.initiate()'
$ docker exec my-mysqld mysql -e 'CREATE DATABASE momy;'

Create momyfile.json like this:

{
  "src": "mongodb://my-mongod:27017/momy",
  "dist": "mysql://root@my-mysqld:3306/momy",
  "collections": {...}
}

Note: you must change username, password, port, ...etc. to fit your environment.

OK, let's run momy with --import option:

$ docker run \
    --interactive --tty --rm \
    --network my-net \
    --mount type=bind,source=$(pwd),target=/workdir \
    cognitom/momy --import

Everything goes well? Then, stop the container (Ctrl + C). Now you can run it as a daemon:

$ docker run \
    --detach --rm \
    --restart unless-stopped \
    --network my-net \
    --mount type=bind,source=$(pwd),target=/workdir \
    cognitom/momy

For contributors

See dev directory.

License

MIT

This library was originally made by @doubaokun as MongoDB-to-MySQL and rewritten by @cognitom.