retries

Documentation

Index

• func ManageRetries(ctx context.Context, process string, maxAttempts int, maxBackoff time.Duration, ...) error

Functions

func ManageRetries

func ManageRetries(
	ctx context.Context,
	process string,
	maxAttempts int,
	maxBackoff time.Duration,
	fn func() (bool, error),
) error

ManageRetries executes the provided function until it "succeeds" or the maximum number of attempts has been exhausted. The delay between attempts increases exponentially, but is capped by the value of maxBackoff argument. "Success" requires the provided function to return a bool indicating that no retry should be attempted (false). The indication that a retry should or should not be attempted is SEPARATE from whatever error the provided function may return, with the decision to retry or not being made solely on the basis of the bool value. This makes it possible to retry failures after the provided function has done its own internal error handling. It also makes it possible to circumvent retries in the event that the error encountered is unrecoverable.

Example usage:

var client amqp.Client
err := retries.ManageRetries(ctx, "amqp connect", 10, 10 * time.Second,
	func() (bool, error) {
		var e error
		if client, e = amqp.Dial(address); e != nil {
			return true, errors.Wrap(e, "error dialing AMQP server")
		}
		return false, nil
	},
)
// Handle err