The "Developing Modules" documentation states:
Include a minimum of dependencies if possible. If there are
dependencies, document them at the top of the module file, and have
the module raise JSON error messages when the import fails.
When docker_service runs on a remote host without PyYAML it crashes with
ImportError.
This patch raises a JSON error message when import fails, but only if
the PyYAML module is actually used. It's only needed when the
"definition" parameter is given.
Signed-off-by: Stefan Hajnoczi <stefanha@gmail.com>
Reading the entire tar file into memory can result in out-of-memory
conditions such as this traceback:
Traceback (most recent call last):
File "/tmp/ansible_YELTSu/ansible_module_docker_image.py", line 486, in load_image
self.client.load_image(image_data)
File "/usr/local/lib/python2.7/dist-packages/docker/api/image.py", line 147, in load_image
res = self._post(self._url("/images/load"), data=data)
...
File "/usr/lib/python2.7/httplib.py", line 997, in endheaders
self._send_output(message_body)
File "/usr/lib/python2.7/httplib.py", line 848, in _send_output
msg += message_body
MemoryError
Luckily docker-py's load_image(), which calls requests post(), accepts a
file-like object instead of a string. Pass in the file object to avoid
reading the full file into memory. This allows larger tar files to load
succesfully.
Signed-off-by: Stefan Hajnoczi <stefanha@gmail.com>
* This moves the lines in the code that parse the `env` and `env_file` options for docker to the end of the `__init__()` function.
This is needed because the `_check_capabilites` function needs both a working `self.client` and a proper `self.docker_py_versioninfo`.
`_check_capabilities` is used by `ensure_capabilities` which is, in turn, used by `get_environment`
This means that before this commit, the environment variables could not be loaded because both `self.client` and `self.docker_py_versioninfo` were not set at that time.
This commit fixes that by putting the environment variable parsing after those two.
* This moves the lines in the code that parse the `env` and `env_file` options for docker to the end of the `__init__()` function.
This is needed because the `_check_capabilites` function needs both a working `self.client` and a proper `self.docker_py_versioninfo`.
`_check_capabilities` is used by `ensure_capabilities` which is, in turn, used by `get_environment`
This means that before this commit, the environment variables could not be loaded because both `self.client` and `self.docker_py_versioninfo` were not set at that time.
This commit fixes that by putting the environment variable parsing after those two.
* Adding docker_container
* If state absent, stop the container before attempting to remove. Fixed status running check.
* If container absent, stop before removing. Fix container status check.
Apologies, but I no longer use this module day-to-day myself, and I don't have the bandwidth right now to effectively triage changes in any kind of timely fashion.
Hello!
I wanted stop the containers matched only by image name, but can't do this, if I not set cmd in playbook.
This behavior confused me.
If cmd or entrypoint is defined for running container, but not defined in playbook, makes matching behavior as this sample:
https://github.com/ansible/ansible-modules-core/blob/devel/cloud/docker/docker.py#L463
restart_containers(containers.running) may try to restart containers
that are deleted when looping through get_differing_containers()
fix this by refreshing list after first loop
The ulimit will be specified as a list and separated by colons. The
hard limit is optional, in which case it is equal to the soft limit.
The ulimits are compared to the ulimits of the container and added
or adjusted accordingly on by a reload.
The module ensures that ulimits are available in the capabilities
iff ulimits is passes as a parameter.
Previously the logging module hard coded the default logging driver. This means
if the docker daemon is started with a different logging driver, the ansible
module would continually restart it when run.
This fix adds a call to docker.Client.info(), which is inspected if a logging
driver is not supplied in the playbook, and the container only restarted if
the logging driver applied differs from the configured default.
In usage, this has solved issues with using alternative logging drivers.
Since we now have several exceptions to the assumption that the
result of the pull would be on the last status line returned by
docker-py's pull(), I've changed the function so that it looks
through the status lines and returns what if finds on it.
Despite the repeated `break`s, the code seems simpler and a little
more coherent like this. From what I've checked using
`https://github.com/jlafon/ansible-profile`, the execution time is
mostly the same.
Before this patch:
- Command was matched if 'Command' field of docker-py
representation of Docker container ends with 'command' passed
to Ansible docker module by user.
- That can give false positives and false negatives.
- For example:
a) If 'command' was set up with more than one spaces,
like 'command=sleep 123', it would be never matched again
with a container(s) launched by this task.
Because after launching, command would be normalized and
appear, in docker-py API call, just as 'sleep 123' - with one
space. This is false negative case.
b) If 'entrypoint + command = command', for example
'sleep + 123 = sleep 123', module would give false positive
match.
This patch fixes it, by making matching more explicit - against
'Config'->Cmd' field of 'docker inspect' output, provided by docker-py
API and with proper normalization of user input by splitting it to
tokens with 'shlex.split()'.
Give user a course of action in the case where the suggestions do not
work. This will hopefully allow us to work through any further issues
much faster.
Check commit enables using tls when using the docker_image module. It
also removes the default for docker_url which doesn't allow us to check
for DOCKER_HOST which is a more sane default. This allows you to use
docker_image on OSX but more documentation is needed.
When pulling an image using Docker 1.8, it seems the output
JSON stream has an empty dict at the very end. This causes
ansible to fail when pulling an image, as it's expecting a
status message in that dict which it uses to determine whether
it had to download the image or not. As a bit of an ugly hack
for that which remains backward compatible, try the last item
in the stream, and if it's an empty dict, take the last-but-one
item instead.
The strip() is needed as the exact value appears to be '{}/r/n';
we could just match that, but it seems like the kind of thing
where maybe it'd happen to just be '{}/n' or '{}' or something
in some cases, so let's just use strip() in case.
A recent change [1] in docker between v1.8.2 and v1.8.3 changed what
is returned in the json when inspecting an image. Five variables which
could have been expected before will now be omited when empty. Only
one of those variables is being addressed in the docker, ExposedPorts.
Unfortunately there was also no API version change on this so this
can't be easily corrected with pinning the API to the older version.
This does a get() which will return None if the variable is not in the
dict formed from the json that was returned. Everything else works the
same way.
[1] 9098628b29
chouseknecht
Julien Vey
Chris Houseknecht
jjshoe
Chris Houseknecht
Qi Luo
Rodríguez, Jorge
Hagen Kuehn
Stefan Hajnoczi
Vince v. Oosten
pic
Mike Mars
Jirayut 'Dear' Nimsaeng
Dan Schmidt
Brian Coca
Shubham
Matt Martz
Brian Coca
Toshio Kuratomi
John R Barker
Darek Kaczyński
ThomasSteinbach
Alexey Wasilyev
Fabian von Feilitzsch
Ash Wilson
Michael Scherer
Victor Beresnev
Yannig Perré
Schlueter
Andrey Gubarev
Tobias Gesellchen
Konstantin Suvorov
Daan Oosterveld
Alexey Wasilyev
jefferson fermo
Will Rouesnel
Pavel Derendyaev
sebi
Manuel Tiago Pereira
Andrew Pashkin
Omar Khan
Joe Quadrino
Yann Hamon
Steve Spencer
Leonty
Aaron Boushley
Philippe Jandot
Michael Schuett
Adam Williamson
Sam Yaple
Colin Hutchinson
justnom
David Dyball
Omri Iluz
Ilya Epifanov