Programming/Error/Handling/Retry Logic/Exponential Backoff

• Exponential Backoff

  • A retry strategy where the delay between retry attempts increases exponentially

  • Key Concepts

    • Initial delay: The first waiting period after a failure
    • Multiplier/Base: The factor by which the delay increases each time (typically 2)
    • Max delay: Upper limit on how long to wait between retries
    • Max attempts: Maximum number of retry attempts before giving up

  • Common Use Cases

    • Network requests and API calls
    • Distributed systems communication
    • Rate limiting and throttling
    • Database connection retries

  • Benefits

    • Prevents overwhelming systems under stress
    • Allows temporary issues to resolve naturally
    • Reduces network congestion
    • More efficient than fixed-interval retries

  • Example Formula

    • delay = min(max_delay, initial_delay * (base ^ attempt_number))
    • For base=2:
      • 1st retry: 1s
      • 2nd retry: 2s
      • 3rd retry: 4s
      • 4th retry: 8s
      • etc.
    • Note:
      • The key property of exponential backoff is multiplicative growth of the wait-time:

        
        Displayed: t_n = t_0 \times f^{\,n},
        
        where Inline: f>1 is a constant factor and Inline: n is the retry count.
        Doubling (Inline: f = 2) yields the classic binary-exponential sequence
        
        Displayed: 1\text{ s} \;\to\; 2\text{ s} \;\to\; 4\text{ s} \;\to\; 8\text{ s} \;\dots
        
        which is what the code does with delay = initial_delay * (base ^ attempt_number).
        Multiplying by a fixed constant on each failure therefore produces the exponential curve Inline: t_n = t_0\,2^{\,n}.
        

  • Implementation Examples

    • LangChain/Blog/25/04/17 LangChain Python Improved Content Blocks Retry Logic and More - LangChain's Runnable.with_retry implementation

  • See also

    • LangChain/output_parsers/retry/RetryOutputParser