Row-level security for embedded viewers

For example, let's assume developer X has a web application for their clients. Clients log in to the app using their own credentials and can view sales data for different stores. Each client has access to a different list of stores. The developer would like to embed a Looker Studio dashboard in their application so that when a client logs in, the client sees only the sales data for the stores they have access to. Clients should not have to log into their Google account for this process to work.

The proposed solution will need the developer to create a Community Connector, pass a token through the embed URL, and then filter data based on the token.

Requirements

  • Dashboard viewers will log in to the third-party application.
  • The application should pass a unique token to the Looker Studio dashboard through embed URL. The token can either be used to lookup filter information or have the filter information encrypted within.
  • The Community Connector should be able to convert the token into the filter values.

Limitations

  • If you are a G Suite customer and your administrator has disabled sharing Drive files to “Anyone with a Link”, you will not be able to share the report with users outside your organization.

Solution

Complete all the following steps to implement the solution.

Generate user token in web app

Generate a unique token for each logged-in user in your web application.You will be passing this token to the embedded dashboard in a later step.

You should use the token to filter relevant data. Options include:

  • Create an API endpoint which return filtered data or the user information for a specific token.
  • Encrypt the user information in the token so that it can be decrypted later in the connector.

Create a new Community Connector

Review How Community Connectors Work and complete the Community Connector Codelab to get started. Use the Local development tooling for creating connectors for a faster and easier development process.

Write the connector code

  1. getConfig() should return at least one config item. This will be used to capture a token from embed URL's parameters.

    function getConfig(request) {
  var cc = DataStudioApp.createCommunityConnector();
  var config = cc.getConfig();

  config
      .newTextInput()
      .setId('token')
      .setName('Enter user token');

  // TODO: Add additional config values if applicable for your connector

  config.setDateRangeRequired(false);
  config.setIsSteppedConfig(false);

  return config.build();
}

  2. getData() will have access to the token through the request.configParams object. Use the token to fetch filtered data or filter existing fetched data.
    Following the above example, the token would be request.configParams.token. In getData(), the token is passed to a REST API end-point to get a list of Store Ids. These Store Ids are then used to construct the SQL query to fetch sales data.

    var STORE_ID_API = 'https://www.example.com/api/storeid';
var QUERY_STRING_PREFIX = "SELECT StoreName, Sales from stores"

function getData(request) {
  var token = request.configParams.token;

  var storeIds = getStoreIdList(token);
  var queryString = constructQueryString(storeIds);
  var fetchedData = fetchData(queryString);

  // rest of getData() implementation
}

function getStoreIdList(token) {
  var url = STORE_ID_API;
  var response = UrlFetchApp.fetch(url);
  var parsedResponse = JSON.parse(response);
  return parsedResponse.storeIds;
}

function constructQueryString(storeIds) {
  var storeIdString = storeIds.join(',');
  var queryString = QUERY_STRING_PREFIX
                    + ' WHERE storeId in ('
                    + storeIdString
                    + ')';
  return queryString;
}

Create the dashboard

  1. Understand how deployments and versions work for connectors.
  2. Create a production deployment for the connector.
  3. Use the Production deployment link to create a data source and a report in Looker Studio.
  4. For the token config parameter, Allow report viewers to modify parameter values.
  5. Share the dashboard with selected users or with “Anyone with link”.
  6. Enable embedding for the report.

Embed the dashboard in your platform

  1. Understand how Report URL parameters work.
  2. Pass the dynamically generated token value using URL parameters to the embedded Looker Studio report.
    Your embed url will look similar to this:
    `https://lookerstudio.google.com/embed/reporting/REPORT_ID/page/PAGE_ID?config=%7B%22ds0%22%3A%7B%22token%22%3A%22TOKEN_VALUE%22%7D%7D

Recommendations

  • You should create a short-lived token.
  • Ensure the dashboard does not leak any information with viewed with an invalid token.