CODEWITHSUNDEEP
node.jssequelize.js
MonkeyBonkey
MonkeyBonkey

Reputation: 47911

How to programmatically run sequelize migrations

The documentation for sequelize seems out of date as they no longer support running migrations from sequelize itself, but instead relies on sequelize-cli. Is there an example of how to use sequeliz-cli programmatically to run the latest migrations? All the documentation seems to be focused on using the client in a shell.

db.js seems to have the function db:migrate that perhaps I can include.

https://github.com/sequelize/cli/blob/master/lib/tasks/db.js

Upvotes: 17

Views: 17385

Answers (3)

broofa
broofa

Reputation: 38151

Update: @phil-court's solution is better.

Original answer here for posterity...

I dug into the code for the sequelize db:migrate command, and there's enough going on there that, IMHO, the simplest/best approach is to just run the command in a child process. Here's the code I used for this (as an await'ed Promise):

const {exec} = require('child_process');

await new Promise((resolve, reject) => {
  const migrate = exec(
    'sequelize db:migrate',
    {env: process.env},
    err => (err ? reject(err): resolve())
  );

  // Forward stdout+stderr to this process
  migrate.stdout.pipe(process.stdout);
  migrate.stderr.pipe(process.stderr);
});

Upvotes: 23

Phil Court
Phil Court

Reputation: 311

I had this exact same problem and implemented the accepted answer. However I ran into concurrency issues while running this as a separate process, especially during tests.

I think this question is rather old, but it still appears very high on search results. Today it's a much better idea to run it using umzug. It's the library that sequelize uses to manage migrations on it's end, and is suggested by the docs.

const fs = require('fs');
const Umzug = require('umzug');
const path = require('path');
const Sequelize = require('sequelize');
const { sequelize } = require('../models/index.js');

const umzug = new Umzug({
  migrations: {
    // indicates the folder containing the migration .js files
    path: path.join(process.cwd(), './migrations'),
    // inject sequelize's QueryInterface in the migrations
    params: [
      sequelize.getQueryInterface(),
      Sequelize,
    ],
  },
  // indicates that the migration data should be store in the database
  // itself through sequelize. The default configuration creates a table
  // named `SequelizeMeta`.
  storage: 'sequelize',
  storageOptions: {
    sequelize,
  },
});

async function migrate() {
  return umzug.up();
}

async function revert() {
  return umzug.down({ to: 0 });

And with that you can do everything you need to do with migrations without resorting to spawning a different process, which opens you to all sorts of race conditions and problems down the line. Read more about how to use umzug with the docs on github

Upvotes: 20

TheBlueAssasin
TheBlueAssasin

Reputation: 93

This is what I did. It is not yet extensively tested, and probably can be optimized further:

const Sequelize = require('sequelize');
const db = new Sequelize('main', 'test', 'test', {
dialect: 'sqlite',
// SQLite only
storage: 'db.db'
});

async function checkForMigrations() {
let migrations = fs.readdirSync(__dirname + '/../migrations');
let completedMigrations = await db.query("SELECT * FROM `SequelizeMeta`", {type: Sequelize.QueryTypes.SELECT});
for (let name in completedMigrations) {
    if (completedMigrations.hasOwnProperty(name)) {
        let index = migrations.indexOf(completedMigrations[name].name);
        if (index !== -1) {
            migrations.splice(index, 1);
        }
    }
}

for(let i = 0, c = migrations.length; i < c; i++){
   let migration = require(__dirname + '/../migrations/' + migrations[i]);
   migration.up(db.queryInterface, Sequelize);
   await db.query("INSERT INTO `SequelizeMeta` VALUES(:name)", {type: Sequelize.QueryTypes.INSERT, replacements: {name: migrations[i]}})
}
}

Upvotes: 2

Related Questions