Retry code execution with a random delay between attempts

try_again() runs code multiple times, with a random delay between attempts. The delay is drawn from a uniform distribution on [delay_seconds_min, delay_seconds_max] before every retry. It does not grow with the attempt number, so this is not exponential backoff. Use it for transient failures in operations that a later attempt can succeed at, such as network requests or file operations.

Usage

try_again(
  x,
  times = 2,
  delay_seconds_min = 5,
  delay_seconds_max = 10,
  verbose = FALSE
)

Arguments

x

The code to run, as an expression.

times

Integer. The maximum number of attempts to make. The default is 2.

delay_seconds_min

Numeric. The minimum delay in seconds between attempts. The default is 5.

delay_seconds_max

Numeric. The maximum delay in seconds between attempts. The default is 10.

verbose

Logical. Whether to show progress information. The default is FALSE.

Value

TRUE, invisibly, after an attempt succeeds. try_again() throws an error with the last error message when every attempt fails.

Details

try_again() is adapted from the try_again function in the testthat package. It adds features that control the retry behavior and the verbosity.

See also

vignette("plnr") for an introduction to the framework. try_again() is a general-purpose retry helper. It is not part of the Plan workflow.

Examples

# A call that succeeds on the first attempt returns immediately
print(try_again(1 + 1))
#> [1] TRUE

# A call that fails once and then succeeds on the second attempt. The delays
# are set to zero so the example runs instantly; they default to 5-10 seconds.
attempt_n <- 0
try_again(
  {
    attempt_n <- attempt_n + 1
    if (attempt_n < 2) stop("not ready yet")
    "succeeded"
  },
  times = 3,
  delay_seconds_min = 0,
  delay_seconds_max = 0
)

# The expression really was evaluated twice
attempt_n
#> [1] 2