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
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
|
I've [[/ansible-and-chef/][already write]] about
[[http://www.ansibleworks.com/][Ansible]]. I use it to manage and
configure my servers (most of them are VMs running on my laptop, but the
idea is the same). One of the server is used to store my personal git
repositories, and I wanted to use [[http://git.zx2c4.com/cgit/][cgit]]
as the front end for the public repositories instead of the old and slow
[[https://git.wiki.kernel.org/index.php/Gitweb][gitweb]].
Since there's no package in Debian for cgit, I need to have an easy
procedure to install it. I'll show how I do it with Ansible. This could
be useful if you're learning about Ansible are you're looking for a
simple use case.
** Work directory
The work directory contains a bunch of files:
- $workdir/hosts - local inventory with all the hosts, grouped by
categories
- $workdir/$hostname.yml - play book for a given host (more about this
file later)
- $workdir/roles/git - directory containing templates, tasks and
handlers for installing cgit
- $workdir/vars/$hostname.yml - contains all the variable needed to
install cgit
#+BEGIN_QUOTE
Replace $hostname with the name of the host you want to use for cgit.
#+END_QUOTE
** Handlers
In my case, cgit is hosted behind Nginx, so first, we need a handler to
restart it after changing Nginx's configuration.
#+BEGIN_EXAMPLE
# roles/git/handlers/main.yml
- name: restart nginx
service: name=nginx state=restarted
#+END_EXAMPLE
** Roles
Now we need to define our role for cgit. The idea is to install the
required packages to be able to build cgit, to create the directories
where we will store our repositories, and actually build cgit.
#+BEGIN_EXAMPLE
# roles/git/tasks/main.yml
- name: Set the directory for public repos
file: path=/srv/git/public
owner=www-data
group=www-data
mode=0770 recurse=yes
state=directory
- name: Set the directory for private repos
file: path=/srv/git/private
owner=www-data
group=www-data
mode=0770
recurse=yes
state=directory
- name: Install necessities for cgit
apt: pkg={{ item }} state=installed
with_items:
- build-essential
- autoconf
- automake
- libtool
- libfcgi-dev
- libssl-dev
- spawn-fcgi
- highlight
- fcgiwrap
- name: Create cgit web directory
file: path=/srv/www/{{ cgit_subdomain }}.{{ domain }}
recurse=yes
state=directory
owner=www-data
- name: Download cgit tarbal
get_url: url=http://git.zx2c4.com/cgit/snapshot/cgit-0.9.2.zip
dest=/tmp/cgit-0.9.2.zip
force=no
- name: Unzip cgit
command: unzip -qo /tmp/cgit-0.9.2.zip -d /tmp
- name: Configure cgit installation
template: src=cgit.conf.j2 dest=/tmp/cgit-0.9.2/cgit.conf
- name: Install cgit
shell: make get-git && make && make install chdir=/tmp/cgit-0.9.2
- name: Set permissions for cgit
file: path=/srv/www/{{ cgit_subdomain }}.{{ domain }}
owner=www-data
state=directory
recurse=yes
- name: Configure the nginx HTTP server for cgit
template: src=etc_nginx_sites-available_cgit.j2
dest=/etc/nginx/sites-available/{{ cgit_subdomain }}.{{ domain }}
group=www-data
owner=www-data
- name: Configure cgit
template: src=etc_cgitrc.j2
dest=/etc/cgitrc
group=www-data
owner=www-data
- name: Enable cgit
file: src=/etc/nginx/sites-available/{{ cgit_subdomain }}.{{ domain }}
dest=/etc/nginx/sites-enabled/{{ cgit_subdomain }}.{{ domain }}
state=link
group=www-data
owner=www-data
notify: restart nginx
- name: Backup git directory
template: src=etc_cron.hourly_git-backup.j2
dest=/etc/cron.hourly/git-backup
mode=0755
#+END_EXAMPLE
** Templates
We need a bunch of templates to configure and build our tools. Let's
start with *cgit.conf*.
#+BEGIN_SRC sh
# roles/git/templates/cgit.conf.j2
CGIT_SCRIPT_PATH = /srv/www/{{ cgit_subdomain }}.{{ domain }}
#+END_SRC
This file is used when we build cgit to install it to a specific
location.
The next template is to configure cgit.
#+BEGIN_EXAMPLE
# roles/git/templates/etc_cgitrc.j2
root-desc=Franck Cuny's projects
virtual-root=/
logo=/cgit.png
css=/cgit.css
scan-path=/srv/git/public
remove-suffix=1
clone-prefix=http://git.$hostname.net
#+END_EXAMPLE
This template is to configure nginx.
#+BEGIN_EXAMPLE
# roles/git/templates/etc_nginx_sites-available_cgit.j2
server {
listen 80;
server_name "{{ cgit_subdomain}}.{{ domain }}";
root /srv/www/{{ cgit_subdomain }}.{{ domain }};
location / {
try_files $uri @cgit;
}
location @cgit {
index cgit.cgi;
fastcgi_param SCRIPT_FILENAME $document_root/cgit.cgi;
fastcgi_pass unix:/run/fcgiwrap.socket;
fastcgi_param HTTP_HOST $server_name;
fastcgi_param PATH_INFO $uri;
fastcgi_param QUERY_INFO $uri;
include "fastcgi_params";
}
error_log /var/log/nginx/{{ cgit_subdomain }}.{{ domain }}-error.log;
access_log /var/log/nginx/{{ cgit_subdomain }}.{{ domain }}-access.log;
}
#+END_EXAMPLE
** Backing up on s3
I backup all my git repositories to a bucket on s3. In order to do that,
you'll need either a new role or to update the current one by adding the
following instructions.
#+BEGIN_EXAMPLE
- name: Install s3cmd
apt: pkg=s3cmd
- name: Configure s3cmd
sudo: false
template:
src="s3cfg.j2"
dest="/root/.s3cfg"
- name: Backup git directory
template: src=etc_cron.hourly_git-backup.j2
dest=/etc/cron.hourly/git-backup
mode=0755
#+END_EXAMPLE
We need a template to configure our access to s3.
#+BEGIN_EXAMPLE
[default]
access_key = {{ aws_access_key }}
secret_key = {{ aws_secret_key }}
use_https = True
#+END_EXAMPLE
And another template for our cron job.
#+BEGIN_SRC sh
#!/bin/sh
s3cmd sync -v /srv/git/ s3://$hostname-backup/git/ > /tmp/s3_backup_git.log 2>&1
#+END_SRC
** Variables
I have a file named *vars/$hostname.yml* that contains the
#+BEGIN_EXAMPLE
domain: $hostname.net
cgit_subdomain: git
aws_access_key: access-key
aws_secret_key: secret-key
#+END_EXAMPLE
** Play time
The content of the playbook
#+BEGIN_EXAMPLE
- hosts: $hostname
vars_files:
- vars/$hostname.yml
roles:
- git
#+END_EXAMPLE
Now I can tell Ansible to run this playbook, and this will install cgit
on my server: =ansible-playbook -i hosts lj.yml=.
|
