UnicodeEncodeError

[Eugene Romero]

Hi all,

Trying to run a playbook here, but getting a Unicode error. Here's what I'm doing:

The plays:

main.yml:

- name: Check if new upstart job is needed

  template:

    src=syslog-ng-upstart.conf.j2

    dest=/etc/init/syslog-ng.conf

    owner=root

    group=root

    mode=0644

    backup=yes

  register: upstart

- name: Update upstart configuration

  include: upstart.yml

  when: upstart.changed​​

------------------------

upstart.yml:

- name: Stop the syslog-ng service before upgrading to upstart

  service: name=syslog-ng state=stopped

  

- name: Move old sysV config

  command: mv /etc/init.d/syslog-ng /etc/init.d/syslog-ng.oldconf

  ignore_errors: yes

- name: Link the upstart-job script to the syslog-ng job

  file:

    src=/lib/init/upstart-job

    path=/etc/init.d/syslog-ng

    state=link

    

- name: Reload initctl config

  command: sudo initctl reload-configuration

- name: Change perms on lock folder

  file:

    path=/var/lib/syslog-ng/

    owner=syslog

    group=syslog

    mode=0755

    state=directory

  

- name: Restart syslog-ng service

  service: name=syslog-ng state=started

----------------------

Not sure if important, but the template file syslog-ng-upstart.conf.j2 is:

start on runlevel [2345]

stop on runlevel [!2345]

 

nice 0

 

respawn

respawn limit 3 15

 

normal exit 0

 

console log

 

# defaults

env OPT_ARGS=''

env SYSLOG_GID='syslog'

env SYSLOG_UID='syslog'

env THREADS_NUM=5

 

script

        # load global defaults file

        if [ -f /etc/default/syslog-ng ]; then

                . /etc/default/syslog-ng

        fi

        exec /usr/sbin/syslog-ng ${OPT_ARGS} --foreground --process-mode foreground --stderr --worker-threads "${THREADS_NUM}" --user "${SYSLOG_UID}" --group "${SYSLOG_GID}"

end script

-------------------------------------

When I run this job, it errors out with the following message (server names edited out):

GATHERING FACTS ***************************************************************

ok: [******]

TASK: [syslog-client | Check if new upstart job is needed] ********************

changed: [******]

TASK: [syslog-client | Stop the syslog-ng service before upgrading to upstart] ***

fatal: [******] => Traceback (most recent call last):

  File "/usr/lib/python2.7/dist-packages/ansible/runner/__init__.py", line 561, in _executor

    exec_rc = self._executor_internal(host, new_stdin)

  File "/usr/lib/python2.7/dist-packages/ansible/runner/__init__.py", line 666, in _executor_internal

    return self._executor_internal_inner(host, self.module_name, self.module_args, inject, port, complex_args=complex_args)

  File "/usr/lib/python2.7/dist-packages/ansible/runner/__init__.py", line 756, in _executor_internal_inner

    if not utils.check_conditional(cond, self.basedir, inject, fail_on_undefined=self.error_on_undefined_vars):

  File "/usr/lib/python2.7/dist-packages/ansible/utils/__init__.py", line 255, in check_conditional

    original = str(conditional).replace("jinja2_compare ","")

UnicodeEncodeError: 'ascii' codec can't encode characters in position 15-16: ordinal not in range(128)

FATAL: all hosts have already failed -- aborting

I have commented out the play that it errors out on, but the error just happens on the next play. So I think the error might have to do with the " register: upstart" bit. But I can't figure out where the problem is.

Thanks in advance for any help you can provide.

[Eugene Romero]

Also, I tried changing "module_lang" to "en_US.UTF-8" in /etc/ansible/ansible.cfg, but it made no difference. I don't have a ansible.cfg in the playbook directory or .ansible.cfg in the home directory.

Ansible version is 1.7.2.

Thanks!

[Toshio Kuratomi]

There have been sporadic UnicodeErrors that crop up from time to time.
I've been trying squash them as they come up. 1.7.2 is pretty old so
it won't have a lot of those fixes. Are you able to test (at least
the failing task) on a newer version of ansible? That would tell you
if the problem has been fixed already in a newer version.

If it's still failing with the newer version, narrowing down the
problem would be helpful. Is it occurring with the first task or the
second? (I'm guessing the first). What is the output from
ansible-playbook -vvvv ? What LC* and LANG settings do you have on the
remote box?

-Toshio

> --
> You received this message because you are subscribed to the Google Groups
> "Ansible Project" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to ansible-proje...@googlegroups.com.
> To post to this group, send email to ansible...@googlegroups.com.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/ansible-project/f788e29d-2c2a-40a1-91b9-cd55e85b41ea%40googlegroups.com.
>
> For more options, visit https://groups.google.com/d/optout.

[Eugene Romero]

Hey Toshio! Thanks for your reply.

Since these are production machines, I can only use 1.7.2 at the moment. However, I installed the newest ppa version (1.9.2) on a VM cluster to test.

The error stills pops up. The error happens when it tries running the first task from the second playbook. However, commenting that task out just makes the error pop up in the next task. So I believe it's already carrying it from the first playbook, after completing the "Check if new upstart job is needed" task. 

The LC and LANG settings are "en_US.UTF-8" in the remote box.

-vvvv doesn't give me much new info. Here's the compressed output when the error occurs:

ok: [anakin] => {"changed": false, "gid": 0, "group": "root", "mode": "0644", "owner": "root", "path": "/etc/init/syslog-ng.conf", "size": 615, "state": "file", "uid": 0}

TASK: [syslog-client | debug var=upstart] *******************************

<anakin> ESTABLISH CONNECTION FOR USER: ******

ok: [anakin] => {

    "var": {

        "upstart_times": {

            "changed": false,

            "gid": 0,

            "group": "root",

            "invocation": {

                "module_args": "src=syslog-ng-upstart.conf.j2 dest=/etc/init/syslog-ng.conf owner=root group=root mode=0644 backup=yes",

                "module_name": "template"

            },

            "mode": "0644",

            "owner": "root",

            "path": "/etc/init/syslog-ng.conf",

            "size": 615,

            "state": "file",

            "uid": 0

        }

    }

}

TASK: [syslog-client | Stop the syslog-ng service before upgrading to upstart] ***

Traceback (most recent call last):

  File "/usr/bin/ansible-playbook", line 324, in <module>

    sys.exit(main(sys.argv[1:]))

  File "/usr/bin/ansible-playbook", line 264, in main

    pb.run()

  File "/usr/lib/pymodules/python2.7/ansible/playbook/__init__.py", line 348, in run

    if not self._run_play(play):

  File "/usr/lib/pymodules/python2.7/ansible/playbook/__init__.py", line 789, in _run_play

    if not self._run_task(play, task, False):

  File "/usr/lib/pymodules/python2.7/ansible/playbook/__init__.py", line 497, in _run_task

    results = self._run_task_internal(task, include_failed=include_failed)

  File "/usr/lib/pymodules/python2.7/ansible/playbook/__init__.py", line 439, in _run_task_internal

    results = runner.run()

  File "/usr/lib/pymodules/python2.7/ansible/runner/__init__.py", line 1493, in run

    results = [ self._executor(h, None) for h in hosts ]

  File "/usr/lib/pymodules/python2.7/ansible/runner/__init__.py", line 590, in _executor

    msg = str(ae)

UnicodeEncodeError: 'ascii' codec can't encode characters in position 40-41: ordinal not in range(128)

--------------

I noticed the error happens regardless of the variable.changed being true or false.

Let me know what other info you might need.

Thanks a lot for your help!

[Toshio Kuratomi]

Not a lot to go on... I do see that the two errors are happening in
different places. That might be bad news for you as may mean that any
fix for 1.9.3 might not be the fix needed for 1.7.2. But let's see
what we can find:

This is an error when trying to format this exception into a string
for display. We can probably just substitute that line like this:

- msg = str(ae)
+ from ansible.utils.unicode import to_bytes
+ msg = to_bytes(ae)

That should solve this traceback but the fact we're getting an
exception there means that something else has gone wrong... It could
be something that you're expected to be able to handle by changing
configuration in your network though, so doing this so that you can
see what the actual error is should be helpful.

If we're lucky, once you see what the underlying error is here, you
can correct something else on your network and then 1.7.2 will run for
you as well.

I'll add a patch to the stable-1.9 tree to do this as well. Not sure
if it will make 1.9.3 or not, though.

https://github.com/ansible/ansible/commit/f80494e434d36c1845c9fb830b6bd26ebcf35a89

>> >> fatal: [******] => Traceback (most recent call last):
>> >> File "/usr/lib/python2.7/dist-packages/ansible/runner/__init__.py",
>> >> line
>> >> 561, in _executor
>> >> exec_rc = self._executor_internal(host, new_stdin)
>> >> File "/usr/lib/python2.7/dist-packages/ansible/runner/__init__.py",
>> >> line
>> >> 666, in _executor_internal
>> >> return self._executor_internal_inner(host, self.module_name,
>> >> self.module_args, inject, port, complex_args=complex_args)
>> >> File "/usr/lib/python2.7/dist-packages/ansible/runner/__init__.py",
>> >> line
>> >> 756, in _executor_internal_inner
>> >> if not utils.check_conditional(cond, self.basedir, inject,
>> >> fail_on_undefined=self.error_on_undefined_vars):
>> >> File "/usr/lib/python2.7/dist-packages/ansible/utils/__init__.py",
>> >> line
>> >> 255, in check_conditional
>> >>

>> >> UnicodeEncodeError: 'ascii' codec can't encode characters in position
>> >> 15-16: ordinal not in range(128)

1.7.2 is old and I'm not sure of some of the ramifications of making
changes in the code (in newer versions of ansible we've been moving
towards making sure that internally we handle everything as unicode
strings and only using bytes when sending things externally. In 1.7.2
we weren't doing that so most things were likely byte strings. jinja2
takes and returns unicode type, however, so that's likely why we have
a problem converting from what jinja2 is returning to us here.) but we
may be able to fix this with the following code change:

- original = str(conditional).replace("jinja2_compare ","")
+ tmp_conditional = to_bytes(conditional,
errors='strict').replace("jinja2_compare ","")
+ original = tmp_conditional.replace("jinja2_compare ","")

Not sure if that will just bring us another error, further down the
line, though.

-Toshio

[Eugene Romero]

Thanks again for your reply Toshio. I apologize for not getting back to you sooner, was off of work for a week and a half. While I was gone, one of my fellow sysadmins updated ansible to 1.9.3, so I am now able to test the failing playbook in production. Once I get up to speed with everything and have a chance to check, I'll post back here if I'm still having issues. Let's hope 1.9.3 fixed the problem :)

Gene