summary refs log tree commit diff
path: root/doc/s6-svscan-not-1.html
blob: 2444e36ef7f9914f4209ccfb5a95b4ad4b8f64e0 (plain) (blame)
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
<html>
  <head>
    <meta name="viewport" content="width=device-width, initial-scale=1.0" />
    <meta http-equiv="Content-Type" content="text/html; charset=UTF-8" />
    <meta http-equiv="Content-Language" content="en" />
    <title>s6: How to run s6-svscan under another init process</title>
    <meta name="Description" content="s6: s6-svscan as not-init" />
    <meta name="Keywords" content="s6 supervision svscan s6-svscan init process boot" />
    <!-- <link rel="stylesheet" type="text/css" href="//skarnet.org/default.css" /> -->
  </head>
<body>

<p>
<a href="index.html">s6</a><br />
<a href="//skarnet.org/software/">Software</a><br />
<a href="//skarnet.org/">skarnet.org</a>
</p>

<h1> How to run s6-svscan under another init process </h1>

<p>
 You can have a reliable supervision tree even if s6-svscan is not your process 1.
The supervision tree just has to be <em>rooted</em> in process 1: that means that
your process 1 will have to supervise your s6-svscan process somehow. That way,
if s6-svscan dies, it will be restarted, and your set of services will always
be maintained.
</p>

<p>
 Be aware, though, that pipes between services and loggers are maintained
by the s6-svscan process; if this process dies, the pipes will be closed and
some logs may be lost.
</p>

<a name="log">
<h2> Logging the supervision tree's output </h2>
</a>

<p>
 s6-svscan and the various s6-supervise processes might produce error or
warning messages; those messages are written to s6-svscan's stderr (which
is inherited by the s6-supervise processes). To log these messages:
</p>

<ul>
 <li> You can use your init system's logging tools, and make your init
system launch s6-svscan as is; its stderr should already be taken care
of by the logging tools. </li>
 <li> You can use a trick similar to the <a href="s6-svscan-1.html#log">process
1 output logging trick</a> so the supervision tree's messages are logged via
a service that's maintained by the supervision tree itself. Then your init
system should not launch s6-svscan directly, but a wrapper script that performs
the proper redirections. The
<tt>examples/s6-svscanboot</tt> file in the s6 distribution gives an example of
such a script. Make sure that your initial <a href="scandir.html">scan
directory</a> contains a <a href="servicedir.html">service directory</a> for your
initial logging service, that must read on the logging FIFO. </li>
</ul>

<p>
 In some of the following examples, we'll assume that <tt>/command/s6-svscanboot</tt>
is the name of the script you are using to start s6-svscan. We will also assume
that all of the s6 executables are available through the <tt>/command</tt> path.
Adjust this accordingly.
</p>

<a name="sysv">
<h2> System V init </h2>
</a>

<p>
 Put an appropriate line in your <tt>/etc/inittab</tt> file, then reload this
config file with <tt>telinit q</tt>.
</p>

<h3> Example </h3>

<pre> SV:123456:respawn:/command/s6-svscanboot </pre>


<a name="upstart">
<h2> Upstart </h2>
</a>

<p>
 Put an appropriate configuration file in the <tt>/etc/init</tt> folder,
for instance <tt>/etc/init/s6-svscan.conf</tt>, then start the service
with <tt>start s6-svscan</tt>.
</p>

<h3> Example </h3>

<pre># s6-svscan
start on runlevel [2345]
stop on runlevel [!2345]

oom never
respawn
exec /command/s6-svscanboot
</pre>

<a name="systemd">
<h2> systemd </h2>
</a>

<p>
 Put an appropriate unit file in the <tt>/etc/systemd/system</tt> folder,
for instance <tt>/etc/systemd/system/s6.service</tt>. It will be picked
up by systemd at boot time.
</p>

<h3> Example </h3>

<pre>[Unit]
Description=s6 supervision tree
Documentation=https://skarnet.org/software/s6/

[Install]
WantedBy=multi-user.target

[Service]
Type=simple
ExecStart=/command/s6-svscan /run/service
ExecReload=/command/s6-svscanctl -an /run/service
Restart=always
RestartSec=1
KillMode=mixed
</pre>

<p>
 Please note that, among other things, systemd performs process supervision,
so depending on the level of integration with your distribution that you
wish to achieve, you may be better off using systemd to directly manage
your daemons. Please also note that systemd is a
<a href="//skarnet.org/software/systemd.html">terrible piece of software engineering</a>, and
if at all possible, you should try and switch to a distribution that does
not use it.
</p>

<a name="bsd">
<h2> BSD init </h2>
</a>

<p>
 Put an appropriate line in your <tt>/etc/ttys</tt> file, then reload this
file with <tt>kill -s HUP 1</tt>.
</p>

<h3> Example </h3>

<pre> sv /command/s6-svscanboot "" on </pre>

<a name="launchd">
<h2> MacOS launchd </h2>
</a>

<p>
 Like systemd, launchd comes with its own
<a href="https://developer.apple.com/library/mac/documentation/macosx/conceptual/bpsystemstartup/chapters/CreatingLaunchdJobs.html">way
of supervising services</a>; if you are a launchd user, you probably do
not need s6.
</p>


</body>
</html>