How to fix authentication errors in Windsor.ai

When connecting a data source to Windsor.ai, users may encounter authentication errors that prevent successful integration. 

These errors typically fall into the following categories:

• Incorrect login credentials
• Insufficient permissions
• Invalid or missing OAuth scopes
  

This guide provides troubleshooting steps for these issues.

Possible reasons:

1. Incorrect username/password.
2. Expired session or outdated stored credentials.
3. Connectivity issues with the data source’s server.
4. Multiple account logins.
5. Browser cache issues.

Troubleshooting steps:

1. Verify credentials – Ensure you are using the correct login details. To avoid typos, copy and paste your credentials instead of typing them manually.
2. Check for service outages – Confirm that the data source’s API or platform (e.g., Google Ads, Facebook Ads) is operational.
3. Refresh or reauthorize the connection – A quick and effective way to resolve these errors is to refresh the link between the data source and the platform. 
4. Check multi-account logins – Windsor.ai can encounter issues during the data integration when you’re logged into more than one Google account. Managing multiple Google accounts properly can help avoid conflicts. You can log out of other accounts or create separate profiles for each of your Google accounts.
5. Clear your browser’s cache and cookies – Deleting your browser’s cache and cookies can eliminate outdated permissions that may interfere with authentication in Windsor.ai. Check your browser’s support guide for detailed instructions on how to do this.

Possible reasons:

1. The connected account lacks the necessary access rights.
2. The data source requires admin-level permissions.
3. Read/write access restrictions are preventing the connection.

Troubleshooting steps:

1. Use an admin account – Ensure the connected account has admin privileges for the data source.
2. Grant required permissions – Check the data source’s settings to allow read/write access.
3. Check role-based access – Ensure the connected user role allows API access.

Possible reasons:

1. OAuth authorization did not request all required scopes.
2. The user revoked access to Windsor.ai from the data source.

Troubleshooting steps:

1. Reauthorize with full scopes – Disconnect and reconnect the data source, ensuring all required permissions are granted.
2. Check OAuth permissions – Visit the data source’s security settings to confirm that Windsor.ai has the right access.
3. Enable additional scopes – If the API requires additional scopes, manually add them before reconnecting.

When setting up a destination in Windsor.ai (e.g., Google BigQuery, Snowflake, Amazon Redshift, Google Sheets, etc.), users may face authentication issues that prevent data from being successfully delivered.

The most common authentication errors include:

• Incorrect credentials (e.g., invalid API key, username, password).
• Insufficient permissions (e.g., missing write access).
• Incorrect Windsor.ai API usage.
• Error in Looker Studio “Bad Request Error 400”

This guide provides troubleshooting steps and solutions for these issues.

Possible reasons:

1. Incorrect database credentials (username/password).
2. API key or URL is invalid.
3. Connection settings (host, port, server) are incorrect.

Troubleshooting steps:

1. Verify credentials – Double-check that the username, password, user and name of the database or table of destination is correct.
2. Check API URL – Check if the Windsor.ai API URL you are using is accessible and returning the required data.
3. Check connection details – Ensure the host, port, server name, and other required parameters are correct.
4. Test connection manually – Try connecting to the destination outside of Windsor.ai using a database client or API request.

Possible reasons:

1. The connected user or service account does not have the required write access.
2. The destination’s security policies block external integrations.

Troubleshooting steps:

1. Grant necessary permissions – Update role-based access in the destination platform.
2. Use a service account (if required) – Some destinations, like BigQuery, require a service account instead of a personal login.
3. Check firewall & IP whitelisting – Some cloud databases (e.g., Snowflake, Redshift) may block connections from unknown IPs. Add the Windsor.ai IP address in the destination firewall settings of the destination platform.

Possible reasons:

1. The API key or URL is invalid.
2. Your Windsor.ai account is expired.

Troubleshooting steps:

1. Check API – If you use Windsor.ai API to get data at your destination, double-check that the API URL and key are correct.
2. Check Windsor.ai account – Ensure that you have an active Windsor.ai subscription and that your destination connections are not out of the limit.

Looker Studio users may encounter such errors as “too many redirects” or a “Bad Request Error 400” when installing the Windsor.ai Looker Studio connector.

Possible reasons:

• Both problems stem from the same root cause: being logged in to multiple Google accounts in your Google Chrome session. This can cause conflicts and lead to these errors, preventing you from successfully installing the connector.

Troubleshooting steps:

1. Log in to only one Google Account. Here’s a detailed guide on how to do this properly: https://windsor.ai/documentation/how-to-fix-the-error-in-looker-studio-bad-request-error-400/.

When connecting a data source to Windsor.ai, users may also encounter authentication errors related to API tokens. These issues can disrupt data retrieval and integration, requiring troubleshooting steps to restore connectivity.

Common authentication errors include:

• Incorrect or missing API tokens
• Expired API tokens
• Insufficient API scopes or permissions
• Invalid credentials

This guide provides solutions to address these issues.

Possible reasons:

1. The API token is not provided in the request.
2. The token format is incorrect or missing the required parameters.
3. Required API scopes are not granted during authentication.

Troubleshooting steps:

1. Verify API token – Ensure that you have a valid and active API token.
2. Check the token format – Confirm that the token follows the correct format specified by the data source.
3. Validate API scopes – Ensure that all required scopes are enabled and granted.
4. Reauthorize the connection – Disconnect and reconnect the data source while granting all necessary permissions.

Possible reasons:

1. API tokens have a limited lifespan and need periodic renewal.
2. The system is using an old or revoked token.
3. The token refresh mechanism is not properly configured.

Troubleshooting steps:

1. Generate a new token – Obtain a new API token from the data source and update the connection.
2. Enable auto-renewal – If the data source supports it, configure automatic token renewal.
3. Store tokens securely – Ensure tokens are stored correctly and not hardcoded in scripts.
4. Re-authenticate the data source – Manually disconnect and reconnect the data source to refresh the token.

Possible reasons:

1. The provided API token or credentials are incorrect.
2. API access is restricted due to insufficient permissions.
3. The API key is associated with an incorrect project or account.

Troubleshooting steps:

1. Double-check credentials – Ensure that the API token, client ID, and secret key are entered correctly.
2. Verify account permissions – Confirm that the connected account has the necessary access rights.
3. Check API key restrictions – Some APIs restrict access based on IP addresses or project settings.
4. Use an admin account – If necessary, switch to an admin account for full API access.
5. Review API documentation – Ensure that the correct authentication flow and token scopes are followed.

By following these troubleshooting steps, users can resolve common authentication errors and ensure a stable connection between Windsor.ai and their data sources and destinations.

If the issue still persists, contact our customer support for further assistance.