Node.js multitenancy guide

This guide will walk you through implementing search in a multitenant Node.js application handling sensitive medical data.

What is multitenancy?

In Meilisearch, you might have one index containing data belonging to many distinct tenants. In such cases, your tenants must only be able to search through their own documents. You can implement this using tenant tokens.

Requirements

Node.js and a package manager like npm, yarn, or pnpm
Meilisearch JavaScript SDK
A Meilisearch server running — see our quick start
A search API key — available in your Meilisearch dashboard
A search API key UID — retrieve it using the keys endpoints

Prefer self-hosting? Read our installation guide.

Data models

This guide uses a simple data model to represent medical appointments. The documents in the Meilisearch index will look like this:

[
  {
    "id": 1,
    "patient": "John",
    "details": "I think I caught a cold. Can you help me?",
    "status": "pending"
  },
  {
    "id": 2,
    "patient": "Zia",
    "details": "I'm suffering from fever. I need an appointment ASAP.",
    "status": "pending"
  },
  {
    "id": 3,
    "patient": "Kevin",
    "details": "Some confidential information Kevin has shared.",
    "status": "confirmed"
  }
]

For the purpose of this guide, we assume documents are stored in an appointments index.

Creating a tenant token

The first step is generating a tenant token that will allow a given patient to search only for their appointments. To achieve this, you must first create a tenant token that filters results based on the patient’s ID. Create a search.js file and use the following code to generate a tenant token:

// search.js

import { Meilisearch } from 'meilisearch'

const apiKey = 'YOUR_SEARCH_API_KEY'
const apiKeyUid = 'YOUR_SEARCH_API_KEY_UID'
const indexName = 'appointments'

const client = new Meilisearch({
  host: 'https://edge.meilisearch.com', // Your Meilisearch host
  apiKey: apiKey
})

export function createTenantToken(patientName) {
  const searchRules = {
    [indexName]: {
      'filter': `patient = ${patientName}`
    }
  }

  const tenantToken = client.generateTenantToken(
    apiKeyUid,
    searchRules,
    {
      expiresAt: new Date('2030-01-01'), // Choose an expiration date
      apiKey: apiKey,
    }
  )
  return tenantToken
}

When Meilisearch gets a search query with a tenant token, it decodes it and applies the search rules to the search request. In this example, the results are filtered by the patient field. This means that a patient can only search for their own appointments.

Using the tenant token

Now that you have a tenant token, use it to perform searches. To achieve this, you will need to:

On the server: create an endpoint to send the token to your front-end
On the client: retrieve the token and use it to perform searches

Serving the tenant token

This guide uses Express.js to create the server. You can install express by running:

# with NPM
npm i express
# with Yarn
yarn add express
# with pnpm
pnpm add express

Then, add the following code in a server.js file:

// server.js

import express from 'express'
import { createTenantToken } from './search.js'

const server = express()

server.get('/token', async (request, response) => {
  const { id: patientId } = request.query
  const token = createTenantToken(patientId)
  return response.json({ token });
})

server.listen(3000, () => {
  console.log('Server is running on port 3000')
})

This code creates an endpoint at http://localhost:3000/token that accepts an id query parameter and returns a tenant token.

Making a search

Now that we have an endpoint, you will use it to retrieve the tenant token in your front-end application. This guide uses InstantSearch.js to create a search interface. You will use CDN links to include InstantSearch.js and the Meilisearch InstantSearch.js connector in your HTML file. Create client.html file and insert this code:

<!DOCTYPE html>
<html lang="en">
  <head>
    <meta charset="utf-8" />
    <link rel="stylesheet" href="https://cdn.jsdelivr.net/npm/@meilisearch/instant-meilisearch/templates/basic_search.css" />
  </head>
  <body>
    <div class="wrapper">
      <div id="searchbox" focus></div>
      <div id="hits"></div>
    </div>
  </body>
  <script src="https://cdn.jsdelivr.net/npm/@meilisearch/instant-meilisearch/dist/instant-meilisearch.umd.min.js"></script>
  <script src="https://cdn.jsdelivr.net/npm/instantsearch.js@4"></script>
  <script>
    document.addEventListener('DOMContentLoaded', async () => {
      const patientId = 1 // Replace with the patient's ID
      const response = await fetch(`http://localhost:3000/token?id=${patientId}`)
      const { token } = await response.json()

      const search = instantsearch({
        indexName: 'appointments',
        searchClient: instantMeiliSearch(
          'https://edge.meilisearch.com',
          token
        ).searchClient
      })

      search.addWidgets([
        instantsearch.widgets.searchBox({
          container: "#searchbox"
        }),
        instantsearch.widgets.hits({
          container: "#hits",
          templates: {
          item: `
            <div>
              <div class="hit-name">
                    {{#helpers.highlight}}{ "attribute": "patient" }{{/helpers.highlight}}
              </div>
              <div class="hit-description">
                {{#helpers.highlight}}{ "attribute": "details" }{{/helpers.highlight}}
              </div>
            </div>
          `
          }
        })
      ])

      search.start()
    })
  </script>
</html>

Ta-da! You have successfully implemented a secure, multitenant search in your Node.js application. Users will only be able to search for documents that belong to them.

Conclusion

In this guide, you saw how to implement secure, multitenant search in a Node.js application. You then created an endpoint to generate tenant tokens for each user. You also built a search interface with InstantSearch to make searches using the tenant token. All the code in this guide is a taken from our multitenacy example application. The code is available on GitHub.

Front end

Back end

Artificial intelligence

Performance

Security

Deployment

Integrations

Miscellaneous

What is multitenancy?

Requirements

Data models

Creating a tenant token

Using the tenant token

Serving the tenant token

Making a search

Conclusion

​What is multitenancy?

​Requirements

​Data models

​Creating a tenant token

​Using the tenant token

​Serving the tenant token

​Making a search

​Conclusion

What is multitenancy?

Requirements

Data models

Creating a tenant token

Using the tenant token

Serving the tenant token

Making a search

Conclusion