-
Notifications
You must be signed in to change notification settings - Fork 0
/
README
261 lines (194 loc) · 9.44 KB
/
README
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
253
254
255
256
257
258
259
260
261
|\_
B A C K U P N I N J A /()/
`\|
a silent flower blossom death strike to lost data.
Backupninja allows you to coordinate system backup by dropping a few
simple configuration files into /etc/backup.d/. Most programs you
might use for making backups don't have their own configuration file
format. Backupninja provides a centralized way to configure and
coordinate many different backup utilities.
Features:
- easy to read ini style configuration files.
- you can drop in scripts to handle new types of backups.
- backup actions can be scheduled
- you can choose when status report emails are mailed to you
(always, on warning, on error, never).
- console-based wizard (ninjahelper) makes it easy to create
backup action configuration files.
- passwords are never sent via the command line to helper programs.
- works with Linux-Vservers (http://linux-vserver.org/)
Backup types:
- secure, remote, incremental filesytem backup (via rdiff-backup).
incremental data is compressed. permissions are retained even
with an unpriviledged backup user.
- backup of mysql databases (via mysqlhotcopy and mysqldump).
- basic system and hardware info
- encrypted remote backups (via duplicity).
- backup of subversion repositories.
The following options are available:
-h, --help This usage message
-d, --debug Run in debug mode, where all log messages are
output to the current shell.
-f, --conffile FILE Use FILE for the main configuration instead
of /etc/backupninja.conf
-t, --test Test run mode. This will test if the backup could run, without actually
preforming any backups. For example, it will attempt to authenticate
or test that ssh keys are set correctly.
-n, --now Perform actions now, instead of when they might be scheduled.
No output will be created unless also run with -d.
--run FILE Runs the specified action FILE (e.g. one of the /etc/backup.d/ files).
Also puts backupninja in debug mode.
NINJAHELPER
===========
Ninjahelper is an additional script which will walk you through the process of
configuring backupninja. Ninjahelper has a menu driven curses based interface
(using dialog).
To add an additional 'wizard' to ninjahelper, follow these steps:
(1) to add a helper for the handler "blue", create the file
blue.helper in the directory where the handlers live.
(ie /usr/share/backupninja).
(2) next, you need to add your helper to the global HELPERS variable
and define the main function for your helper (the function name
is always <helper>_wizard). for example, blue.helper:
HELPERS="$HELPERS blue:description_of_this_helper"
blue_wizard() {
... do work here ...
}
(3) look at the existing helpers to see how they are written. Try to re-use
functions, such as the dialog functions that are defined in easydialog.sh,
or the vserver functions defined in lib/vserver.
(4) test, re-test, and test again. Try to break the helper by going backwards,
try to think like someone who has no idea how to configure your handler
would think, try to make your helper as simple as possible. Walk like a cat,
become your shadow, don't let your senses betray you.
CONFIGURATION FILES
===================
The general configuration file is /etc/backupninja.conf. In this file
you can set the log level and change the default directory locations.
You can force a different general configuration file with "backupninja
-f /path/to/conf".
To preform the actual backup, backupninja processes each configuration
file in /etc/backup.d according to the file's suffix:
.sh -- run this file as a shell script.
.rdiff -- filesystem backup (using rdiff-backup)
.dup -- filesystem backup (using duplicity)
.mysql -- backup mysql databases
.pgsql -- backup PostgreSQL databases
.sys -- general hardware, partition, and system reports.
.svn -- backup subversion repositories
.maildir -- incrementally backup maildirs (very specialized)
Support for additional configuration types can be added by dropping
bash scripts with the name of the suffix into /usr/share/backupninja.
The configuration files are processed in alphabetical order. However,
it is suggested that you name the config files in "sysvinit style."
For example:
00-disabled.pgsql
10-runthisfirst.sh
20-runthisnext.mysql
90-runthislast.rdiff
Typically, you will put a '.rdiff' config file last, so that any
database dumps you make are included in the filesystem backup.
Configurations files with names beginning with 0 (zero) or ending with
.disabled (preferred method) are skipped.
Unless otherwise specified, the config file format is "ini style."
For example:
# this is a comment
[fishes]
fish = red
fish = blue
[fruit]
apple = yes
pear = no thanks \
i will not have a pear.
SCHEDULING
==========
By default, each configuration file is processed everyday at 01:00 (1
AM). This can be changed by specifying the 'when' option in a config
file.
For example:
when = sundays at 02:00
when = 30th at 22
when = 30 at 22:00
when = everyday at 01 <-- the default
when = Tuesday at 05:00
A configuration file will be processed at the time(s) specified by the
"when" option. If multiple "when" options are present, then they all
apply. If two configurations files are scheduled to run in the same
hour, then we fall back on the alphabetical ordering specified above.
If two configurations files are scheduled close to one another in
time, it is possible to have multiple copies of backupninja running if
the first instance is not finished before the next one starts.
Make sure that you put the "when" option before any sections in your
configuration file.
These values for 'when' are equivalent:
when = tuesday at 05:30
when = TUESDAYS at 05
These values for 'when' are invalid:
when = tuesday at 2am
when = tuesday at 2
when = tues at 02
REAL WORLD USAGE
================
Backupninja can be used to implement whatever backup strategy you
choose. It is intended, however, to be used like so:
(1) First, databases are safely copied or exported to /var/backups.
Typically, you cannot make a file backup of a database while it
is in use, hence the need to use special tools to make a safe copy
or export into /var/backups.
(2) Then, vital parts of the file system, including /var/backups, are
nightly pushed to a remote, off-site, hard disk (using
rdiff-backup). The local user is root, but the remote user is not
priviledged. Hopefully, the remote filesystem is encrypted.
There are many different backup strategies out there, including "pull
style", magnetic tape, rsync + hard links, etc. We believe that the
strategy outlined above is the way to go because: (1) hard disks are
very cheap these days, (2) pull style backups are no good, because then
the backup server must have root on the production server, and (3)
rdiff-backup is more space efficient and featureful than using rsync +
hard links.
SSH KEYS
========
In order for rdiff-backup to sync files over ssh unattended, you must
create ssh keys on the source server and copy the public key to the
remote user's authorized keys file. For example:
root@srchost# ssh-keygen -t rsa -b 4096
root@srchost# ssh-copy-id -i /root/.ssh/id_rsa.pub backup@desthost
Now, you should be able to ssh from user 'root' on srchost to
user 'backup' on desthost without specifying a password.
Note: when prompted for a password by ssh-keygen, just leave it
blank by hitting return.
The included helper program "ninjahelper" will walk you through creating
an rdiff-backup configuration, and will set up the ssh keys for you.
Amazon Simple Storage Service (S3)
==================================
Duplicity can store backups on Amazon S3 buckets, taking care of encryption.
Since it performs incremental backups it minimizes the number of request per
operation therefore reducing the costs. The boto Python interface to Amazon
Web Services is needed to use duplicity with S3 (Debian package: python-boto).
VSERVERS
========
If you are using Linux-Vservers (http://linux-vserver.org/) there are some
special capabilities that different handlers have to make vserver
backups easier.
Set the variable "vservers" to be "yes" in /etc/backupninja.conf and see the
example configuration files for each handler to configure the vserver specific
variables.
Additional vserver variables that can be configured in /etc/backupninja.conf,
but they probably don't need to be changed:
VSERVERINFO (default: /usr/sbin/vserver-info)
VSERVER (default: /usr/sbin/vserver)
VROOTDIR (default: `$VSERVERINFO info SYSINFO |grep vserver-Rootdir | awk '{print $2}'`)
.sh CONFIGURATION FILES
=======================
Shell jobs may use the following features:
* logging and control flow functions:
halt, fatal, error, warning, info, debug, passthru.
All such functions take a list of strings a parameters.
Those strings are passed to whatever logging mechanism is enabled,
and colored if relevant.
* Using "exit N" is useless, and has unspecified consequences.
Just don't do it.
* when=TIME works as documented above; at may also be written
"when = TIME".
* The $BACKUPNINJA_DEBUG environment variable is set when
backupninja is invoked with the -d option.