Handling Meeting Platform Internal Errors

Occasionally, bots can experience a fatal error due to a dropped connection or server error on the meeting platform's end. This is typically a transient error and not related to your configuration or our service.

This article explains the best practices for handle these errors effectively.

How to Identify these Errors

These errors are indicated by a fatal event containing a sub_code with an _internal_error suffix. Some examples include:

• google_meet_internal_error

• microsoft_teams_internal_error

• zoom_internal_error

How to Handle these Errors

While our system automatically attempts to reconnect the bot several times, there are cases where all retries might fail. In such situations, we recommend the following approach:

1. Implement a retry mechanism in your application to request a new bot when this error occurs.

2. Use an exponential backoff strategy to prevent potential infinite loops in case of persistent issues.

Best Practices for Retry Logic

When implementing your retry logic, consider the following:

• Start with a short delay (e.g., a few seconds) before the first retry.

• Increase the delay exponentially for subsequent retries.

• Set a maximum number of retry attempts to avoid excessive retries during prolonged outages.

• Implement a maximum total retry duration (e.g., 15 minutes) to eventually fail gracefully.

Following these strategies can ensure your application is robust to any potential meeting platform server disruptions.