1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
|
openlmi-scripts
===============
Client-side python modules and command line utilities.
It comprises of one binary called `lmi` and a common library. `lmi`
meta-command allows to run commands on a set of OpenLMI providers. These
commands can be installed separately in a modular way.
`lmi` is a command line application allowing to run single command on a set
of hosts with just one statement from `shell` or it can run in an
interactive way.
For more information please refer to online documentation on [pythonhosted][]
or build your own in `doc/` directory.
Structure
---------
Following diagram depicts directory structure.
openlmi-tools
├── commands # base directory for lmi subcommands
│ ├── service # service provider comand (service)
│ │ └── lmi
│ │ └── scripts
│ │ └── service
│ └── software # software provider command (sw)
│ └── lmi
│ └── scripts
│ └── software
├── config # configuration files for lmi meta-command
└── lmi # common client-side library
└── scripts
├── common
└── _metacommand # functionality of lmi meta-command
Each subdirectory of `commands/` contains library for interfacing with
particular set of OpenLMI providers. Each contains its own `setup.py` file,
that handles its installation and registration of command. They have one
command thing. Each such `setup.py` must pass `entry_points` dictionary to
the `setup()` function, wich associates commands defined in command library
with its name under `lmi` meta-command.
Dependencies
------------
Code base is written for `python 2.7`.
There are following python dependencies:
* openlmi-tools
* python-docopt
### Uploading to PyPI
Since *PyPI* expects README file to be in a *reStructuredText* markup
language and the one present is written in *markdown*, it needs to be
converted to it. So please make sure you have `pandoc` installed before
running:
$ python setup.py sdist upload
Installation
------------
Use standard `setuptools` script for installation:
$ cd openlmi-scripts
$ python setup.py install --user
This installs just the *lmi meta-command* and client-side library. To install
subcommands, you need to do the same procedure for each particular command
under `commands/` directory.
Usage
-----
To get a help and see available commands, run:
$ lmi help
To get a help for particular command, run:
$ lmi help service
To issue single command on a host, run:
$ lmi --host ${hostname} service list
To start the app in interactive mode:
$ lmi --host ${hostname}
> service list --disabled
...
> service start svnserve.service
...
> quit
Developing lmi scripts.
-----------------------
This documents how to quickly develop lmi scripts without the need to
reinstall python eggs, when anything is changed. This presumes, that the
development process takes place in a git repotory checked out from [git][].
It can be located anywhere on system.
Before we start with setting up an environment, please double check, that you
don't have installed anything from openlmi-scripts in system path
(`/usr/lib/python2.7/site-packages/lmi/scripts` should not exist). And make
sure, that user path is also cleared:
$ rm -rf $HOME/.local/lib/python2.7/site-packages/lmi*
$ rm -rf $HOME/.local/lib/python2.7/site-packages/openlmi*
Install all dependencies:
* python-docopt
* openlmi-python-base
* openlmi-tools
Either via rpms or from respective git repositories. For openlmi-python-base
package contained in [providers-git][] repository the setup script is
located at `src/python/setup.py`. In future these will be available from PyPi.
Let's setup an environment:
1. Create a workspace directory for current `$USER` (let's call it a `WSP`).
This is a place, where our eggs and binaries will be "installed".
It can be located anywhere, for example:
$ WSP=~/.python_workspace
$ mkdir $WSP
2. Add workspace to your python path to make all modules installed there
importable (you can add this to your `~/.bashrc`):
$ export PYTHONPATH=$WSP:$PYTHONPATH
3. Add workspace to your PATH, so the installed binaries can be run:
$ export PATH=$WSP:$PATH
4. Now let's "install" to our workspace. First `cd` to checked out
openlmi-scripts repository.
5. Install them and any commands you want -- possibly your own
$ python setup.py develop --install-dir=$WSP
$ for cmd in service storage; do
> pushd commands/$cmd
> python setup.py develop --install-dir=$WSP
> popd
> done
Now any change made to openlmi-scripts is immediately reflected in `lmi`
meta-command.
------------------------------------------------------------------------------
[git]: https://github.com/openlmi/openlmi-scripts "openlmi-scripts"
[providers-git]: https://fedorahosted.org/openlmi/browser/openlmi-providers "openlmi-providers"
[pythonhosted]: http://pythonhosted.org/openlmi-scripts/index.html "python hosted"
|