'None' timeout breaks with Client.stop

[ruiztulio]

Hi,

We are wotking on a script to automate some stuff using docker and docker-py in some point we execute some code inside docker, this command can take from a few minutes to a 5 or 6 hours.

I realized that I could do something like this:

container_name = 'container_name_or_id'
cli = Client(timeout=None)
output = cli.execute(container_name, 'ls')

And works as expected, but then the container have to be stopped and started (here I just put the specific code for a PoC):

from docker import Client

container_name = 'container_name_or_id'
cli = Client(timeout=None)

# Here goes a command that take a lot of time, not just a ls
output = cli.execute(container_name, 'ls')
cli.stop(container_name)
# Do some more stuff
cli.start(container_name)

But we get this:

Traceback (most recent call last):
  File "test_resume.py", line 48, in <module>
    cli.stop(container_name)
  File "/home/truiz/deployer/venv/lib/python2.7/site-packages/docker/client.py", line 923, in stop
    timeout=(timeout + self.timeout))
TypeError: unsupported operand type(s) for +: 'int' and 'NoneType'

Is there any other way of setting an "infinite" timeout? or just no timeout for this long runnig scripts?

Regards

[shin-]

Hm, that's definitely a bug. Thank you for reporting it.