Fix docker shell 2

[msosyak]

Description of PR

In this PR #1253 I have tested docker plugin by running dip_sib and lag2 (lldp task run in lldp container) test cases and everything was fine.
But after regression, I have noticed that some tests failed due to this plugin with error:
KeyError: 'Requested entry (plugin_type: shell plugin: docker setting: ansible_python_interpreter ) was not defined in configuration.' on copy task.

Here is the fix for the error described above.

Summary:
Fixed docker shell plugin part 2
Fixes # (issue)

Type of change

• Bug fix
• [] Testbed and Framework(new/improvement)
• [] Test case(new/improvement)

Approach

How did you do it?

Defined mandatory container_name option for this plugin
Change remove and build_module_command to use it as a container name source and replace:

vars:
    ansible_shell_type: docker
    ansible_python_interpreter: docker exec -i swss python

to

vars:
    ansible_shell_type: docker
    container_name: swss

How did you verify/test it?

run ecn_wred, coop and dip_sib test_cases

Any platform specific information?

Supported testbed topology if it's a new test case?

Documentation

[msosyak]

@qiluo-msft could you have a look at this?

[qiluo-msft]

I suspect you are not fixing the root issue. I did a while and found one missing in docker.py, you may need to override ShellBase.append_command().

Please help open an issue, and provide the verbose error message in order to better communicate.

[msosyak]

Related issue #1268
@qiluo-msft I am not sure that append_command() will really help us as it just called to concatenate two commands. I go deep and have made this PR #1269 where I used native docker connection plugin.

[msosyak]

Closed In favor of #1269