Skip to main content

A lightweight and pythonic retry helper

Project description

Build Status Coverage Status


A lightweight and pythonic retry helper.

tryagain aims to simplify working with unstable functions. Whether you have networking code that sometimes raises timeout exceptions or you are controlling devices which only seem to listen on the second try - tryagain makes it easier to repeat the call.

tryagain offers you hooks to clean up after a failed attempt or to prepare for the next call. You can set a waittime between retries or specify your own waittime function to realize exponential waittimes etc.

tryagain is lightweight, fully tested, MIT licensed and comes as a single python file with no dependencies. It supports Python 2.6+ and 3.2+.

To install, run pip install tryagain.

Basic syntax

Using the tryagain function call:

import tryagain

def unstable_function():
    # Attention: This function sometimes fails!

result =,
                       max_attempts=None, exceptions=Exception, wait=0.0,
                       cleanup_hook=None, pre_retry_hook=None)

Using the tryagain decorator retries:

from tryagain import retries

def unstable_funcation(arg1, arg2):
    # Attention: This function sometimes fails!

result = unstable_function('foo', arg2='bar')


  • func: The unstable function to call

  • max_attemps: Any integer number to limit the maximum number of attempts. Set to None for unlimited retries. (Default = None)

  • exceptions: An iterable of exceptions that should result in a retry. (Default = Exception)

  • wait: Can be an integer or float value (to specify a waittime in seconds) or a custom function (see Waittime documentation) (Default = 0.0)

  • cleanup_hook: Can be set to a callable and will be called after an exception is raised from calling func. (Default = None)

  • pre_retry_hook: Can be set to any callable that will be called before func is called. (Default = None)

Result will return whatever the unstable function would return. (and the decorator tryagain.retries) reraises any exception which is:

  • not in the given exceptions

  • raised in the pre_retry_hook or in cleanup_hook

  • raised in the last attempt at calling the unstable function.


Retry calling an unstable function

import tryagain

def unstable():

# retry calling 'unstable' until it returns without raising an exception
result =

# limit to maximum 5 attempts
result =, max_attempts=5)

# only retry after specific exceptions
result =, exceptions=(ValueError, TypeError))


The tryagain library allows fixed wait values as well as custom waittime functions.

# wait one second before trying again, wait=1.0)

# waittime rises linearly (n is the number of attempts)
# (will wait 1s, 2s, 3s, ...), wait=lambda n: n)

# waittime rises exponentially with each attempt
# (will wait 2s, 4s, 8s, ...), wait=lambda n: 2 ** n)

# exponentially rising waittime with maximum
# (will wait 2s, 4s, 5s, 5s, ..., 5s), wait=lambda n: min(n ** 2, 5))

# no waiting time before second attempt, 1.0s afterwards
def no_first_wait(attempt):
    if attempt == 2:
        return 0
        return 1.0, wait=no_first_wait)

Retry calling a function with parameters

The only supports a function reference as the func parameter. To pass arguments to the unstable function you have to use one of the following idioms:

# using a lambda unstable('message', some_arg=True), wait=1.0)

# using a partial
from functools import partial, 'message', some_arg=True), wait=1.0)

# using a separate function
def call_unstable_function():
    msg = 'message'
    return unstable(msg, some_arg=True), wait=1.0)

But it is much nicer to wrap your unstable function in the @retries decorator. This way you can call your unstable function with parameters easily:

Function decorator

Instead of using the function, you can use the retries decorator.

from tryagain import retries
@retries(max_attempts=3, exceptions=(TypeError, ValueError))
def unstable(arg1, arg2):
    # your unstable function here

result = unstable('foo', arg2='bar')

The decorator takes the same arguments as the call-function except the func parameter.


The tryagain library features two hooks that can be used, cleanup_hook and pre_retry_hook.

def unstable():
    print('Calling unstable function')
    raise Exception, max_attempts=2,
              wait=lambda n: print('waiting'),
              cleanup_hook=lambda: print('cleaning up'),
              pre_retry_hook=lambda: print('do preparations'))
'Calling unstable function'
'cleaning up'
'do preparations'
'Calling unstable function'
'cleaning up'
Error: Exception raised...

Project details

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distributions

No source distribution files available for this release.See tutorial on generating distribution archives.

Built Distribution

tryagain-1.0-py2.py3-none-any.whl (7.8 kB view hashes)

Uploaded 3 5

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page