]>
Commit | Line | Data |
---|---|---|
6300502b | 1 | <?xml version='1.0'?> <!--*- Mode: nxml; nxml-child-indent: 2; indent-tabs-mode: nil -*--> |
663996b3 | 2 | <!DOCTYPE refentry PUBLIC "-//OASIS//DTD DocBook XML V4.2//EN" |
e735f4d4 | 3 | "http://www.oasis-open.org/docbook/xml/4.2/docbookx.dtd"> |
663996b3 MS |
4 | |
5 | <!-- | |
52ad194e MB |
6 | SPDX-License-Identifier: LGPL-2.1+ |
7 | ||
663996b3 MS |
8 | This file is part of systemd. |
9 | ||
10 | Copyright 2010 Lennart Poettering | |
11 | ||
12 | systemd is free software; you can redistribute it and/or modify it | |
13 | under the terms of the GNU Lesser General Public License as published by | |
14 | the Free Software Foundation; either version 2.1 of the License, or | |
15 | (at your option) any later version. | |
16 | ||
17 | systemd is distributed in the hope that it will be useful, but | |
18 | WITHOUT ANY WARRANTY; without even the implied warranty of | |
19 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU | |
20 | Lesser General Public License for more details. | |
21 | ||
22 | You should have received a copy of the GNU Lesser General Public License | |
23 | along with systemd; If not, see <http://www.gnu.org/licenses/>. | |
24 | --> | |
25 | ||
26 | <refentry id="systemd.socket"> | |
e735f4d4 MP |
27 | <refentryinfo> |
28 | <title>systemd.socket</title> | |
29 | <productname>systemd</productname> | |
30 | ||
31 | <authorgroup> | |
32 | <author> | |
33 | <contrib>Developer</contrib> | |
34 | <firstname>Lennart</firstname> | |
35 | <surname>Poettering</surname> | |
36 | <email>lennart@poettering.net</email> | |
37 | </author> | |
38 | </authorgroup> | |
39 | </refentryinfo> | |
40 | ||
41 | <refmeta> | |
42 | <refentrytitle>systemd.socket</refentrytitle> | |
43 | <manvolnum>5</manvolnum> | |
44 | </refmeta> | |
45 | ||
46 | <refnamediv> | |
47 | <refname>systemd.socket</refname> | |
48 | <refpurpose>Socket unit configuration</refpurpose> | |
49 | </refnamediv> | |
50 | ||
51 | <refsynopsisdiv> | |
52 | <para><filename><replaceable>socket</replaceable>.socket</filename></para> | |
53 | </refsynopsisdiv> | |
54 | ||
55 | <refsect1> | |
56 | <title>Description</title> | |
57 | ||
58 | <para>A unit configuration file whose name ends in | |
59 | <literal>.socket</literal> encodes information about an IPC or | |
60 | network socket or a file system FIFO controlled and supervised by | |
61 | systemd, for socket-based activation.</para> | |
62 | ||
63 | <para>This man page lists the configuration options specific to | |
64 | this unit type. See | |
65 | <citerefentry><refentrytitle>systemd.unit</refentrytitle><manvolnum>5</manvolnum></citerefentry> | |
66 | for the common options of all unit configuration files. The common | |
67 | configuration items are configured in the generic [Unit] and | |
68 | [Install] sections. The socket specific configuration options are | |
69 | configured in the [Socket] section.</para> | |
70 | ||
71 | <para>Additional options are listed in | |
72 | <citerefentry><refentrytitle>systemd.exec</refentrytitle><manvolnum>5</manvolnum></citerefentry>, | |
73 | which define the execution environment the | |
74 | <option>ExecStartPre=</option>, <option>ExecStartPost=</option>, | |
75 | <option>ExecStopPre=</option> and <option>ExecStopPost=</option> | |
76 | commands are executed in, and in | |
77 | <citerefentry><refentrytitle>systemd.kill</refentrytitle><manvolnum>5</manvolnum></citerefentry>, | |
78 | which define the way the processes are terminated, and in | |
79 | <citerefentry><refentrytitle>systemd.resource-control</refentrytitle><manvolnum>5</manvolnum></citerefentry>, | |
80 | which configure resource control settings for the processes of the | |
81 | socket.</para> | |
82 | ||
83 | <para>For each socket file, a matching service file must exist, | |
84 | describing the service to start on incoming traffic on the socket | |
85 | (see | |
86 | <citerefentry><refentrytitle>systemd.service</refentrytitle><manvolnum>5</manvolnum></citerefentry> | |
87 | for more information about .service files). The name of the | |
88 | .service unit is by default the same as the name of the .socket | |
89 | unit, but can be altered with the <option>Service=</option> option | |
90 | described below. Depending on the setting of the | |
91 | <option>Accept=</option> option described below, this .service | |
92 | unit must either be named like the .socket unit, but with the | |
93 | suffix replaced, unless overridden with <option>Service=</option>; | |
94 | or it must be a template unit named the same way. Example: a | |
95 | socket file <filename>foo.socket</filename> needs a matching | |
96 | service <filename>foo.service</filename> if | |
97 | <option>Accept=false</option> is set. If | |
98 | <option>Accept=true</option> is set, a service template file | |
99 | <filename>foo@.service</filename> must exist from which services | |
100 | are instantiated for each incoming connection.</para> | |
101 | ||
f5e65279 | 102 | <para>No implicit <varname>WantedBy=</varname> or |
e735f4d4 MP |
103 | <varname>RequiredBy=</varname> dependency from the socket to the |
104 | service is added. This means that the service may be started | |
105 | without the socket, in which case it must be able to open sockets | |
106 | by itself. To prevent this, an explicit | |
107 | <varname>Requires=</varname> dependency may be added.</para> | |
108 | ||
109 | <para>Socket units may be used to implement on-demand starting of | |
110 | services, as well as parallelized starting of services. See the | |
111 | blog stories linked at the end for an introduction.</para> | |
112 | ||
113 | <para>Note that the daemon software configured for socket | |
114 | activation with socket units needs to be able to accept sockets | |
115 | from systemd, either via systemd's native socket passing interface | |
116 | (see | |
117 | <citerefentry><refentrytitle>sd_listen_fds</refentrytitle><manvolnum>3</manvolnum></citerefentry> | |
118 | for details) or via the traditional | |
e3bff60a | 119 | <citerefentry project='freebsd'><refentrytitle>inetd</refentrytitle><manvolnum>8</manvolnum></citerefentry>-style |
e735f4d4 MP |
120 | socket passing (i.e. sockets passed in via standard input and |
121 | output, using <varname>StandardInput=socket</varname> in the | |
122 | service file).</para> | |
123 | </refsect1> | |
124 | ||
db2df898 | 125 | <refsect1> |
f5e65279 MB |
126 | <title>Implicit Dependencies</title> |
127 | ||
128 | <para>The following dependencies are implicitly added:</para> | |
129 | ||
130 | <itemizedlist> | |
131 | <listitem><para>Socket units automatically gain a <varname>Before=</varname> | |
132 | dependency on the service units they activate.</para></listitem> | |
133 | ||
134 | <listitem><para>Socket units referring to file system paths (such as AF_UNIX | |
135 | sockets or FIFOs) implicitly gain <varname>Requires=</varname> and | |
136 | <varname>After=</varname> dependencies on all mount units | |
137 | necessary to access those paths.</para></listitem> | |
138 | ||
139 | <listitem><para>Socket units using the <varname>BindToDevice=</varname> | |
140 | setting automatically gain a <varname>BindsTo=</varname> and | |
141 | <varname>After=</varname> dependency on the device unit | |
142 | encapsulating the specified network interface.</para></listitem> | |
143 | </itemizedlist> | |
db2df898 MP |
144 | |
145 | <para>Additional implicit dependencies may be added as result of | |
146 | execution and resource control parameters as documented in | |
147 | <citerefentry><refentrytitle>systemd.exec</refentrytitle><manvolnum>5</manvolnum></citerefentry> | |
148 | and | |
149 | <citerefentry><refentrytitle>systemd.resource-control</refentrytitle><manvolnum>5</manvolnum></citerefentry>.</para> | |
150 | </refsect1> | |
151 | ||
f5e65279 MB |
152 | <refsect1> |
153 | <title>Default Dependencies</title> | |
154 | ||
155 | <para>The following dependencies are added unless | |
156 | <varname>DefaultDependencies=no</varname> is set:</para> | |
157 | ||
158 | <itemizedlist> | |
159 | <listitem><para>Socket units automatically gain a | |
160 | <varname>Before=</varname> dependency on | |
161 | <filename>sockets.target</filename>.</para></listitem> | |
162 | ||
163 | <listitem><para>Socket units automatically gain a pair of | |
164 | <varname>After=</varname> and <varname>Requires=</varname> | |
165 | dependency on <filename>sysinit.target</filename>, and a pair of | |
166 | <varname>Before=</varname> and <varname>Conflicts=</varname> | |
167 | dependencies on <filename>shutdown.target</filename>. These | |
168 | dependencies ensure that the socket unit is started before normal | |
169 | services at boot, and is stopped on shutdown. Only sockets | |
170 | involved with early boot or late system shutdown should disable | |
171 | <varname>DefaultDependencies=</varname> option.</para></listitem> | |
172 | </itemizedlist> | |
173 | </refsect1> | |
174 | ||
e735f4d4 MP |
175 | <refsect1> |
176 | <title>Options</title> | |
177 | ||
178 | <para>Socket files must include a [Socket] section, which carries | |
179 | information about the socket or FIFO it supervises. A number of | |
180 | options that may be used in this section are shared with other | |
181 | unit types. These options are documented in | |
182 | <citerefentry><refentrytitle>systemd.exec</refentrytitle><manvolnum>5</manvolnum></citerefentry> | |
183 | and | |
184 | <citerefentry><refentrytitle>systemd.kill</refentrytitle><manvolnum>5</manvolnum></citerefentry>. | |
185 | The options specific to the [Socket] section of socket units are | |
186 | the following:</para> | |
187 | ||
188 | <variablelist class='unit-directives'> | |
189 | <varlistentry> | |
190 | <term><varname>ListenStream=</varname></term> | |
191 | <term><varname>ListenDatagram=</varname></term> | |
192 | <term><varname>ListenSequentialPacket=</varname></term> | |
193 | <listitem><para>Specifies an address to listen on for a stream | |
194 | (<constant>SOCK_STREAM</constant>), datagram | |
195 | (<constant>SOCK_DGRAM</constant>), or sequential packet | |
196 | (<constant>SOCK_SEQPACKET</constant>) socket, respectively. | |
197 | The address can be written in various formats:</para> | |
198 | ||
199 | <para>If the address starts with a slash | |
200 | (<literal>/</literal>), it is read as file system socket in | |
201 | the <constant>AF_UNIX</constant> socket family.</para> | |
202 | ||
203 | <para>If the address starts with an at symbol | |
204 | (<literal>@</literal>), it is read as abstract namespace | |
205 | socket in the <constant>AF_UNIX</constant> family. The | |
206 | <literal>@</literal> is replaced with a | |
207 | <constant>NUL</constant> character before binding. For | |
208 | details, see | |
209 | <citerefentry project='man-pages'><refentrytitle>unix</refentrytitle><manvolnum>7</manvolnum></citerefentry>.</para> | |
210 | ||
211 | <para>If the address string is a single number, it is read as | |
212 | port number to listen on via IPv6. Depending on the value of | |
213 | <varname>BindIPv6Only=</varname> (see below) this might result | |
214 | in the service being available via both IPv6 and IPv4 | |
215 | (default) or just via IPv6. | |
216 | </para> | |
217 | ||
218 | <para>If the address string is a string in the format | |
219 | v.w.x.y:z, it is read as IPv4 specifier for listening on an | |
220 | address v.w.x.y on a port z.</para> | |
221 | ||
222 | <para>If the address string is a string in the format [x]:y, | |
223 | it is read as IPv6 address x on a port y. Note that this might | |
224 | make the service available via IPv4, too, depending on the | |
225 | <varname>BindIPv6Only=</varname> setting (see below). | |
226 | </para> | |
227 | ||
2897b343 MP |
228 | <para>If the address string is a string in the format |
229 | <literal>vsock:x:y</literal>, it is read as CID <literal>x</literal> on | |
230 | a port <literal>y</literal> address in the | |
231 | <constant>AF_VSOCK</constant> family. The CID is a unique 32-bit | |
232 | integer identifier in <constant>AF_VSOCK</constant> analogous to an IP | |
233 | address. Specifying the CID is optional, and may be set to the empty | |
234 | string.</para> | |
235 | ||
e735f4d4 MP |
236 | <para>Note that <constant>SOCK_SEQPACKET</constant> (i.e. |
237 | <varname>ListenSequentialPacket=</varname>) is only available | |
238 | for <constant>AF_UNIX</constant> sockets. | |
239 | <constant>SOCK_STREAM</constant> (i.e. | |
240 | <varname>ListenStream=</varname>) when used for IP sockets | |
241 | refers to TCP sockets, <constant>SOCK_DGRAM</constant> (i.e. | |
242 | <varname>ListenDatagram=</varname>) to UDP.</para> | |
243 | ||
db2df898 | 244 | <para>These options may be specified more than once, in which |
e735f4d4 MP |
245 | case incoming traffic on any of the sockets will trigger |
246 | service activation, and all listed sockets will be passed to | |
247 | the service, regardless of whether there is incoming traffic | |
248 | on them or not. If the empty string is assigned to any of | |
249 | these options, the list of addresses to listen on is reset, | |
250 | all prior uses of any of these options will have no | |
251 | effect.</para> | |
252 | ||
253 | <para>It is also possible to have more than one socket unit | |
254 | for the same service when using <varname>Service=</varname>, | |
255 | and the service will receive all the sockets configured in all | |
256 | the socket units. Sockets configured in one unit are passed in | |
257 | the order of configuration, but no ordering between socket | |
258 | units is specified.</para> | |
259 | ||
260 | <para>If an IP address is used here, it is often desirable to | |
261 | listen on it before the interface it is configured on is up | |
262 | and running, and even regardless of whether it will be up and | |
263 | running at any point. To deal with this, it is recommended to | |
264 | set the <varname>FreeBind=</varname> option described | |
265 | below.</para></listitem> | |
266 | </varlistentry> | |
267 | ||
268 | <varlistentry> | |
269 | <term><varname>ListenFIFO=</varname></term> | |
270 | <listitem><para>Specifies a file system FIFO to listen on. | |
271 | This expects an absolute file system path as argument. | |
272 | Behavior otherwise is very similar to the | |
273 | <varname>ListenDatagram=</varname> directive | |
274 | above.</para></listitem> | |
275 | </varlistentry> | |
276 | ||
277 | <varlistentry> | |
278 | <term><varname>ListenSpecial=</varname></term> | |
279 | <listitem><para>Specifies a special file in the file system to | |
280 | listen on. This expects an absolute file system path as | |
281 | argument. Behavior otherwise is very similar to the | |
282 | <varname>ListenFIFO=</varname> directive above. Use this to | |
283 | open character device nodes as well as special files in | |
284 | <filename>/proc</filename> and | |
285 | <filename>/sys</filename>.</para></listitem> | |
286 | </varlistentry> | |
287 | ||
288 | <varlistentry> | |
289 | <term><varname>ListenNetlink=</varname></term> | |
290 | <listitem><para>Specifies a Netlink family to create a socket | |
291 | for to listen on. This expects a short string referring to the | |
292 | <constant>AF_NETLINK</constant> family name (such as | |
293 | <varname>audit</varname> or <varname>kobject-uevent</varname>) | |
294 | as argument, optionally suffixed by a whitespace followed by a | |
295 | multicast group integer. Behavior otherwise is very similar to | |
296 | the <varname>ListenDatagram=</varname> directive | |
297 | above.</para></listitem> | |
298 | </varlistentry> | |
299 | ||
300 | <varlistentry> | |
301 | <term><varname>ListenMessageQueue=</varname></term> | |
302 | <listitem><para>Specifies a POSIX message queue name to listen | |
303 | on. This expects a valid message queue name (i.e. beginning | |
304 | with /). Behavior otherwise is very similar to the | |
305 | <varname>ListenFIFO=</varname> directive above. On Linux | |
306 | message queue descriptors are actually file descriptors and | |
307 | can be inherited between processes.</para></listitem> | |
308 | </varlistentry> | |
309 | ||
6300502b MP |
310 | <varlistentry> |
311 | <term><varname>ListenUSBFunction=</varname></term> | |
312 | <listitem><para>Specifies a <ulink | |
313 | url="https://www.kernel.org/doc/Documentation/usb/functionfs.txt">USB | |
8a584da2 | 314 | FunctionFS</ulink> endpoints location to listen on, for |
6300502b | 315 | implementation of USB gadget functions. This expects an |
8a584da2 MP |
316 | absolute file system path of functionfs mount point as the argument. |
317 | Behavior otherwise is very similar to the <varname>ListenFIFO=</varname> | |
db2df898 | 318 | directive above. Use this to open the FunctionFS endpoint |
6300502b MP |
319 | <filename>ep0</filename>. When using this option, the |
320 | activated service has to have the | |
321 | <varname>USBFunctionDescriptors=</varname> and | |
322 | <varname>USBFunctionStrings=</varname> options set. | |
323 | </para></listitem> | |
324 | </varlistentry> | |
325 | ||
4c89c718 MP |
326 | <varlistentry> |
327 | <term><varname>SocketProtocol=</varname></term> | |
328 | <listitem><para>Takes a one of <option>udplite</option> | |
329 | or <option>sctp</option>. Specifies a socket protocol | |
330 | (<constant>IPPROTO_UDPLITE</constant>) UDP-Lite | |
331 | (<constant>IPPROTO_SCTP</constant>) SCTP socket respectively. </para> | |
332 | </listitem> | |
333 | </varlistentry> | |
334 | ||
e735f4d4 MP |
335 | <varlistentry> |
336 | <term><varname>BindIPv6Only=</varname></term> | |
337 | <listitem><para>Takes a one of <option>default</option>, | |
338 | <option>both</option> or <option>ipv6-only</option>. Controls | |
339 | the IPV6_V6ONLY socket option (see | |
e3bff60a | 340 | <citerefentry project='die-net'><refentrytitle>ipv6</refentrytitle><manvolnum>7</manvolnum></citerefentry> |
e735f4d4 MP |
341 | for details). If <option>both</option>, IPv6 sockets bound |
342 | will be accessible via both IPv4 and IPv6. If | |
343 | <option>ipv6-only</option>, they will be accessible via IPv6 | |
344 | only. If <option>default</option> (which is the default, | |
345 | surprise!), the system wide default setting is used, as | |
346 | controlled by | |
347 | <filename>/proc/sys/net/ipv6/bindv6only</filename>, which in | |
348 | turn defaults to the equivalent of | |
349 | <option>both</option>.</para> | |
350 | </listitem> | |
351 | </varlistentry> | |
352 | ||
353 | <varlistentry> | |
354 | <term><varname>Backlog=</varname></term> | |
355 | <listitem><para>Takes an unsigned integer argument. Specifies | |
356 | the number of connections to queue that have not been accepted | |
357 | yet. This setting matters only for stream and sequential | |
358 | packet sockets. See | |
359 | <citerefentry><refentrytitle>listen</refentrytitle><manvolnum>2</manvolnum></citerefentry> | |
360 | for details. Defaults to SOMAXCONN (128).</para></listitem> | |
361 | </varlistentry> | |
362 | ||
363 | <varlistentry> | |
364 | <term><varname>BindToDevice=</varname></term> | |
365 | <listitem><para>Specifies a network interface name to bind | |
366 | this socket to. If set, traffic will only be accepted from the | |
367 | specified network interfaces. This controls the | |
db2df898 MP |
368 | SO_BINDTODEVICE socket option (see <citerefentry |
369 | project='man-pages'><refentrytitle>socket</refentrytitle><manvolnum>7</manvolnum></citerefentry> | |
f5e65279 | 370 | for details). If this option is used, an implicit dependency |
e735f4d4 MP |
371 | from this socket unit on the network interface device unit |
372 | (<citerefentry><refentrytitle>systemd.device</refentrytitle><manvolnum>5</manvolnum></citerefentry> | |
db2df898 MP |
373 | is created. Note that setting this parameter might result in |
374 | additional dependencies to be added to the unit (see | |
375 | above).</para></listitem> | |
e735f4d4 MP |
376 | </varlistentry> |
377 | ||
378 | <varlistentry> | |
379 | <term><varname>SocketUser=</varname></term> | |
380 | <term><varname>SocketGroup=</varname></term> | |
381 | ||
382 | <listitem><para>Takes a UNIX user/group name. When specified, | |
383 | all AF_UNIX sockets and FIFO nodes in the file system are | |
384 | owned by the specified user and group. If unset (the default), | |
385 | the nodes are owned by the root user/group (if run in system | |
386 | context) or the invoking user/group (if run in user context). | |
387 | If only a user is specified but no group, then the group is | |
388 | derived from the user's default group.</para></listitem> | |
389 | </varlistentry> | |
390 | ||
391 | <varlistentry> | |
392 | <term><varname>SocketMode=</varname></term> | |
393 | <listitem><para>If listening on a file system socket or FIFO, | |
394 | this option specifies the file system access mode used when | |
395 | creating the file node. Takes an access mode in octal | |
396 | notation. Defaults to 0666.</para></listitem> | |
397 | </varlistentry> | |
398 | ||
399 | <varlistentry> | |
400 | <term><varname>DirectoryMode=</varname></term> | |
401 | <listitem><para>If listening on a file system socket or FIFO, | |
402 | the parent directories are automatically created if needed. | |
403 | This option specifies the file system access mode used when | |
404 | creating these directories. Takes an access mode in octal | |
405 | notation. Defaults to 0755.</para></listitem> | |
406 | </varlistentry> | |
407 | ||
408 | <varlistentry> | |
409 | <term><varname>Accept=</varname></term> | |
410 | <listitem><para>Takes a boolean argument. If true, a service | |
411 | instance is spawned for each incoming connection and only the | |
412 | connection socket is passed to it. If false, all listening | |
413 | sockets themselves are passed to the started service unit, and | |
414 | only one service unit is spawned for all connections (also see | |
415 | above). This value is ignored for datagram sockets and FIFOs | |
416 | where a single service unit unconditionally handles all | |
417 | incoming traffic. Defaults to <option>false</option>. For | |
418 | performance reasons, it is recommended to write new daemons | |
419 | only in a way that is suitable for | |
420 | <option>Accept=false</option>. A daemon listening on an | |
421 | <constant>AF_UNIX</constant> socket may, but does not need to, | |
422 | call | |
423 | <citerefentry><refentrytitle>close</refentrytitle><manvolnum>2</manvolnum></citerefentry> | |
424 | on the received socket before exiting. However, it must not | |
425 | unlink the socket from a file system. It should not invoke | |
426 | <citerefentry><refentrytitle>shutdown</refentrytitle><manvolnum>2</manvolnum></citerefentry> | |
427 | on sockets it got with <varname>Accept=false</varname>, but it | |
428 | may do so for sockets it got with | |
429 | <varname>Accept=true</varname> set. Setting | |
430 | <varname>Accept=true</varname> is mostly useful to allow | |
431 | daemons designed for usage with | |
e3bff60a | 432 | <citerefentry project='freebsd'><refentrytitle>inetd</refentrytitle><manvolnum>8</manvolnum></citerefentry> |
e735f4d4 | 433 | to work unmodified with systemd socket |
e3bff60a MP |
434 | activation.</para> |
435 | ||
db2df898 MP |
436 | <para>For IPv4 and IPv6 connections, the <varname>REMOTE_ADDR</varname> |
437 | environment variable will contain the remote IP address, and <varname>REMOTE_PORT</varname> | |
e3bff60a | 438 | will contain the remote port. This is the same as the format used by CGI. |
db2df898 | 439 | For SOCK_RAW, the port is the IP protocol.</para></listitem> |
e735f4d4 MP |
440 | </varlistentry> |
441 | ||
6300502b MP |
442 | <varlistentry> |
443 | <term><varname>Writable=</varname></term> | |
444 | <listitem><para>Takes a boolean argument. May only be used in | |
445 | conjunction with <varname>ListenSpecial=</varname>. If true, | |
446 | the specified special file is opened in read-write mode, if | |
db2df898 | 447 | false, in read-only mode. Defaults to false.</para></listitem> |
6300502b MP |
448 | </varlistentry> |
449 | ||
e735f4d4 MP |
450 | <varlistentry> |
451 | <term><varname>MaxConnections=</varname></term> | |
452 | <listitem><para>The maximum number of connections to | |
453 | simultaneously run services instances for, when | |
454 | <option>Accept=true</option> is set. If more concurrent | |
455 | connections are coming in, they will be refused until at least | |
456 | one existing connection is terminated. This setting has no | |
457 | effect on sockets configured with | |
458 | <option>Accept=false</option> or datagram sockets. Defaults to | |
459 | 64.</para></listitem> | |
460 | </varlistentry> | |
461 | ||
462 | <varlistentry> | |
8a584da2 MP |
463 | <term><varname>MaxConnectionsPerSource=</varname></term> |
464 | <listitem><para>The maximum number of connections for a service per source IP address. | |
465 | This is very similar to the <varname>MaxConnections=</varname> directive | |
466 | above. Disabled by default.</para> | |
467 | </listitem> | |
468 | </varlistentry> | |
469 | ||
470 | <varlistentry> | |
e735f4d4 MP |
471 | <term><varname>KeepAlive=</varname></term> |
472 | <listitem><para>Takes a boolean argument. If true, the TCP/IP | |
473 | stack will send a keep alive message after 2h (depending on | |
474 | the configuration of | |
475 | <filename>/proc/sys/net/ipv4/tcp_keepalive_time</filename>) | |
476 | for all TCP streams accepted on this socket. This controls the | |
477 | SO_KEEPALIVE socket option (see | |
e3bff60a | 478 | <citerefentry project='man-pages'><refentrytitle>socket</refentrytitle><manvolnum>7</manvolnum></citerefentry> |
e735f4d4 MP |
479 | and the <ulink |
480 | url="http://www.tldp.org/HOWTO/html_single/TCP-Keepalive-HOWTO/">TCP | |
481 | Keepalive HOWTO</ulink> for details.) Defaults to | |
482 | <option>false</option>.</para></listitem> | |
483 | </varlistentry> | |
484 | ||
485 | <varlistentry> | |
486 | <term><varname>KeepAliveTimeSec=</varname></term> | |
db2df898 | 487 | <listitem><para>Takes time (in seconds) as argument. The connection needs to remain |
e735f4d4 MP |
488 | idle before TCP starts sending keepalive probes. This controls the TCP_KEEPIDLE |
489 | socket option (see | |
e3bff60a | 490 | <citerefentry project='man-pages'><refentrytitle>socket</refentrytitle><manvolnum>7</manvolnum></citerefentry> |
e735f4d4 MP |
491 | and the <ulink |
492 | url="http://www.tldp.org/HOWTO/html_single/TCP-Keepalive-HOWTO/">TCP | |
493 | Keepalive HOWTO</ulink> for details.) | |
494 | Defaults value is 7200 seconds (2 hours).</para></listitem> | |
495 | </varlistentry> | |
496 | ||
497 | <varlistentry> | |
498 | <term><varname>KeepAliveIntervalSec=</varname></term> | |
499 | <listitem><para>Takes time (in seconds) as argument between | |
500 | individual keepalive probes, if the socket option SO_KEEPALIVE | |
db2df898 | 501 | has been set on this socket. This controls |
e735f4d4 | 502 | the TCP_KEEPINTVL socket option (see |
e3bff60a | 503 | <citerefentry project='man-pages'><refentrytitle>socket</refentrytitle><manvolnum>7</manvolnum></citerefentry> |
e735f4d4 MP |
504 | and the <ulink |
505 | url="http://www.tldp.org/HOWTO/html_single/TCP-Keepalive-HOWTO/">TCP | |
506 | Keepalive HOWTO</ulink> for details.) Defaults value is 75 | |
507 | seconds.</para></listitem> | |
508 | </varlistentry> | |
509 | ||
510 | <varlistentry> | |
511 | <term><varname>KeepAliveProbes=</varname></term> | |
db2df898 | 512 | <listitem><para>Takes an integer as argument. It is the number of |
e735f4d4 MP |
513 | unacknowledged probes to send before considering the |
514 | connection dead and notifying the application layer. This | |
515 | controls the TCP_KEEPCNT socket option (see | |
e3bff60a | 516 | <citerefentry project='man-pages'><refentrytitle>socket</refentrytitle><manvolnum>7</manvolnum></citerefentry> |
e735f4d4 MP |
517 | and the <ulink |
518 | url="http://www.tldp.org/HOWTO/html_single/TCP-Keepalive-HOWTO/">TCP | |
519 | Keepalive HOWTO</ulink> for details.) Defaults value is | |
520 | 9.</para></listitem> | |
521 | </varlistentry> | |
522 | ||
523 | <varlistentry> | |
524 | <term><varname>NoDelay=</varname></term> | |
525 | <listitem><para>Takes a boolean argument. TCP Nagle's | |
526 | algorithm works by combining a number of small outgoing | |
527 | messages, and sending them all at once. This controls the | |
528 | TCP_NODELAY socket option (see | |
e3bff60a | 529 | <citerefentry project='die-net'><refentrytitle>tcp</refentrytitle><manvolnum>7</manvolnum></citerefentry> |
e735f4d4 MP |
530 | Defaults to <option>false</option>.</para></listitem> |
531 | </varlistentry> | |
532 | ||
533 | <varlistentry> | |
534 | <term><varname>Priority=</varname></term> | |
535 | <listitem><para>Takes an integer argument controlling the | |
536 | priority for all traffic sent from this socket. This controls | |
537 | the SO_PRIORITY socket option (see | |
e3bff60a | 538 | <citerefentry project='man-pages'><refentrytitle>socket</refentrytitle><manvolnum>7</manvolnum></citerefentry> |
e735f4d4 MP |
539 | for details.).</para></listitem> |
540 | </varlistentry> | |
541 | ||
542 | <varlistentry> | |
543 | <term><varname>DeferAcceptSec=</varname></term> | |
544 | ||
545 | <listitem><para>Takes time (in seconds) as argument. If set, | |
546 | the listening process will be awakened only when data arrives | |
547 | on the socket, and not immediately when connection is | |
548 | established. When this option is set, the | |
549 | <constant>TCP_DEFER_ACCEPT</constant> socket option will be | |
550 | used (see | |
e3bff60a | 551 | <citerefentry project='die-net'><refentrytitle>tcp</refentrytitle><manvolnum>7</manvolnum></citerefentry>), |
e735f4d4 MP |
552 | and the kernel will ignore initial ACK packets without any |
553 | data. The argument specifies the approximate amount of time | |
554 | the kernel should wait for incoming data before falling back | |
8a584da2 | 555 | to the normal behavior of honoring empty ACK packets. This |
e735f4d4 MP |
556 | option is beneficial for protocols where the client sends the |
557 | data first (e.g. HTTP, in contrast to SMTP), because the | |
558 | server process will not be woken up unnecessarily before it | |
559 | can take any action. | |
560 | </para> | |
561 | ||
562 | <para>If the client also uses the | |
563 | <constant>TCP_DEFER_ACCEPT</constant> option, the latency of | |
564 | the initial connection may be reduced, because the kernel will | |
565 | send data in the final packet establishing the connection (the | |
566 | third packet in the "three-way handshake").</para> | |
567 | ||
568 | <para>Disabled by default.</para> | |
569 | </listitem> | |
570 | </varlistentry> | |
571 | ||
572 | <varlistentry> | |
573 | <term><varname>ReceiveBuffer=</varname></term> | |
574 | <term><varname>SendBuffer=</varname></term> | |
575 | <listitem><para>Takes an integer argument controlling the | |
576 | receive or send buffer sizes of this socket, respectively. | |
577 | This controls the SO_RCVBUF and SO_SNDBUF socket options (see | |
e3bff60a | 578 | <citerefentry project='man-pages'><refentrytitle>socket</refentrytitle><manvolnum>7</manvolnum></citerefentry> |
e735f4d4 MP |
579 | for details.). The usual suffixes K, M, G are supported and |
580 | are understood to the base of 1024.</para></listitem> | |
581 | </varlistentry> | |
582 | ||
583 | <varlistentry> | |
584 | <term><varname>IPTOS=</varname></term> | |
585 | <listitem><para>Takes an integer argument controlling the IP | |
586 | Type-Of-Service field for packets generated from this socket. | |
587 | This controls the IP_TOS socket option (see | |
e3bff60a | 588 | <citerefentry project='die-net'><refentrytitle>ip</refentrytitle><manvolnum>7</manvolnum></citerefentry> |
e735f4d4 MP |
589 | for details.). Either a numeric string or one of |
590 | <option>low-delay</option>, <option>throughput</option>, | |
591 | <option>reliability</option> or <option>low-cost</option> may | |
592 | be specified.</para></listitem> | |
593 | </varlistentry> | |
594 | ||
595 | <varlistentry> | |
596 | <term><varname>IPTTL=</varname></term> | |
597 | <listitem><para>Takes an integer argument controlling the IPv4 | |
598 | Time-To-Live/IPv6 Hop-Count field for packets generated from | |
599 | this socket. This sets the IP_TTL/IPV6_UNICAST_HOPS socket | |
600 | options (see | |
e3bff60a | 601 | <citerefentry project='die-net'><refentrytitle>ip</refentrytitle><manvolnum>7</manvolnum></citerefentry> |
e735f4d4 | 602 | and |
e3bff60a | 603 | <citerefentry project='die-net'><refentrytitle>ipv6</refentrytitle><manvolnum>7</manvolnum></citerefentry> |
e735f4d4 MP |
604 | for details.)</para></listitem> |
605 | </varlistentry> | |
606 | ||
607 | <varlistentry> | |
608 | <term><varname>Mark=</varname></term> | |
609 | <listitem><para>Takes an integer value. Controls the firewall | |
610 | mark of packets generated by this socket. This can be used in | |
611 | the firewall logic to filter packets from this socket. This | |
612 | sets the SO_MARK socket option. See | |
e3bff60a | 613 | <citerefentry project='die-net'><refentrytitle>iptables</refentrytitle><manvolnum>8</manvolnum></citerefentry> |
e735f4d4 MP |
614 | for details.</para></listitem> |
615 | </varlistentry> | |
616 | ||
617 | <varlistentry> | |
618 | <term><varname>ReusePort=</varname></term> | |
619 | <listitem><para>Takes a boolean value. If true, allows | |
620 | multiple | |
621 | <citerefentry><refentrytitle>bind</refentrytitle><manvolnum>2</manvolnum></citerefentry>s | |
622 | to this TCP or UDP port. This controls the SO_REUSEPORT socket | |
623 | option. See | |
e3bff60a | 624 | <citerefentry project='man-pages'><refentrytitle>socket</refentrytitle><manvolnum>7</manvolnum></citerefentry> |
e735f4d4 MP |
625 | for details.</para></listitem> |
626 | </varlistentry> | |
627 | ||
628 | <varlistentry> | |
629 | <term><varname>SmackLabel=</varname></term> | |
630 | <term><varname>SmackLabelIPIn=</varname></term> | |
631 | <term><varname>SmackLabelIPOut=</varname></term> | |
632 | <listitem><para>Takes a string value. Controls the extended | |
633 | attributes <literal>security.SMACK64</literal>, | |
634 | <literal>security.SMACK64IPIN</literal> and | |
635 | <literal>security.SMACK64IPOUT</literal>, respectively, i.e. | |
636 | the security label of the FIFO, or the security label for the | |
637 | incoming or outgoing connections of the socket, respectively. | |
638 | See <ulink | |
639 | url="https://www.kernel.org/doc/Documentation/security/Smack.txt">Smack.txt</ulink> | |
640 | for details.</para></listitem> | |
641 | </varlistentry> | |
642 | ||
643 | <varlistentry> | |
644 | <term><varname>SELinuxContextFromNet=</varname></term> | |
645 | <listitem><para>Takes a boolean argument. When true, systemd | |
646 | will attempt to figure out the SELinux label used for the | |
647 | instantiated service from the information handed by the peer | |
648 | over the network. Note that only the security level is used | |
649 | from the information provided by the peer. Other parts of the | |
650 | resulting SELinux context originate from either the target | |
651 | binary that is effectively triggered by socket unit or from | |
652 | the value of the <varname>SELinuxContext=</varname> option. | |
653 | This configuration option only affects sockets with | |
654 | <varname>Accept=</varname> mode set to | |
655 | <literal>true</literal>. Also note that this option is useful | |
656 | only when MLS/MCS SELinux policy is deployed. Defaults to | |
657 | <literal>false</literal>. </para></listitem> | |
658 | </varlistentry> | |
659 | ||
660 | <varlistentry> | |
661 | <term><varname>PipeSize=</varname></term> | |
662 | <listitem><para>Takes a size in bytes. Controls the pipe | |
663 | buffer size of FIFOs configured in this socket unit. See | |
664 | <citerefentry><refentrytitle>fcntl</refentrytitle><manvolnum>2</manvolnum></citerefentry> | |
665 | for details. The usual suffixes K, M, G are supported and are | |
666 | understood to the base of 1024.</para></listitem> | |
667 | </varlistentry> | |
668 | ||
669 | <varlistentry> | |
670 | <term><varname>MessageQueueMaxMessages=</varname>, | |
671 | <varname>MessageQueueMessageSize=</varname></term> | |
672 | <listitem><para>These two settings take integer values and | |
673 | control the mq_maxmsg field or the mq_msgsize field, | |
674 | respectively, when creating the message queue. Note that | |
675 | either none or both of these variables need to be set. See | |
e3bff60a | 676 | <citerefentry project='die-net'><refentrytitle>mq_setattr</refentrytitle><manvolnum>3</manvolnum></citerefentry> |
e735f4d4 MP |
677 | for details.</para></listitem> |
678 | </varlistentry> | |
679 | ||
680 | <varlistentry> | |
681 | <term><varname>FreeBind=</varname></term> | |
682 | <listitem><para>Takes a boolean value. Controls whether the | |
683 | socket can be bound to non-local IP addresses. This is useful | |
684 | to configure sockets listening on specific IP addresses before | |
685 | those IP addresses are successfully configured on a network | |
686 | interface. This sets the IP_FREEBIND socket option. For | |
687 | robustness reasons it is recommended to use this option | |
688 | whenever you bind a socket to a specific IP address. Defaults | |
689 | to <option>false</option>.</para></listitem> | |
690 | </varlistentry> | |
691 | ||
692 | <varlistentry> | |
693 | <term><varname>Transparent=</varname></term> | |
694 | <listitem><para>Takes a boolean value. Controls the | |
695 | IP_TRANSPARENT socket option. Defaults to | |
696 | <option>false</option>.</para></listitem> | |
697 | </varlistentry> | |
698 | ||
699 | <varlistentry> | |
700 | <term><varname>Broadcast=</varname></term> | |
701 | <listitem><para>Takes a boolean value. This controls the | |
702 | SO_BROADCAST socket option, which allows broadcast datagrams | |
703 | to be sent from this socket. Defaults to | |
704 | <option>false</option>.</para></listitem> | |
705 | </varlistentry> | |
706 | ||
707 | <varlistentry> | |
708 | <term><varname>PassCredentials=</varname></term> | |
709 | <listitem><para>Takes a boolean value. This controls the | |
710 | SO_PASSCRED socket option, which allows | |
711 | <constant>AF_UNIX</constant> sockets to receive the | |
712 | credentials of the sending process in an ancillary message. | |
713 | Defaults to <option>false</option>.</para></listitem> | |
714 | </varlistentry> | |
715 | ||
716 | <varlistentry> | |
717 | <term><varname>PassSecurity=</varname></term> | |
718 | <listitem><para>Takes a boolean value. This controls the | |
719 | SO_PASSSEC socket option, which allows | |
720 | <constant>AF_UNIX</constant> sockets to receive the security | |
721 | context of the sending process in an ancillary message. | |
722 | Defaults to <option>false</option>.</para></listitem> | |
723 | </varlistentry> | |
724 | ||
725 | <varlistentry> | |
726 | <term><varname>TCPCongestion=</varname></term> | |
727 | <listitem><para>Takes a string value. Controls the TCP | |
728 | congestion algorithm used by this socket. Should be one of | |
729 | "westwood", "veno", "cubic", "lp" or any other available | |
730 | algorithm supported by the IP stack. This setting applies only | |
731 | to stream sockets.</para></listitem> | |
732 | </varlistentry> | |
733 | ||
734 | <varlistentry> | |
735 | <term><varname>ExecStartPre=</varname></term> | |
736 | <term><varname>ExecStartPost=</varname></term> | |
737 | <listitem><para>Takes one or more command lines, which are | |
738 | executed before or after the listening sockets/FIFOs are | |
739 | created and bound, respectively. The first token of the | |
740 | command line must be an absolute filename, then followed by | |
741 | arguments for the process. Multiple command lines may be | |
742 | specified following the same scheme as used for | |
743 | <varname>ExecStartPre=</varname> of service unit | |
744 | files.</para></listitem> | |
745 | </varlistentry> | |
746 | ||
747 | <varlistentry> | |
748 | <term><varname>ExecStopPre=</varname></term> | |
749 | <term><varname>ExecStopPost=</varname></term> | |
750 | <listitem><para>Additional commands that are executed before | |
751 | or after the listening sockets/FIFOs are closed and removed, | |
752 | respectively. Multiple command lines may be specified | |
753 | following the same scheme as used for | |
754 | <varname>ExecStartPre=</varname> of service unit | |
755 | files.</para></listitem> | |
756 | </varlistentry> | |
757 | ||
758 | <varlistentry> | |
759 | <term><varname>TimeoutSec=</varname></term> | |
760 | <listitem><para>Configures the time to wait for the commands | |
761 | specified in <varname>ExecStartPre=</varname>, | |
762 | <varname>ExecStartPost=</varname>, | |
763 | <varname>ExecStopPre=</varname> and | |
764 | <varname>ExecStopPost=</varname> to finish. If a command does | |
765 | not exit within the configured time, the socket will be | |
766 | considered failed and be shut down again. All commands still | |
767 | running will be terminated forcibly via | |
768 | <constant>SIGTERM</constant>, and after another delay of this | |
769 | time with <constant>SIGKILL</constant>. (See | |
770 | <option>KillMode=</option> in | |
771 | <citerefentry><refentrytitle>systemd.kill</refentrytitle><manvolnum>5</manvolnum></citerefentry>.) | |
772 | Takes a unit-less value in seconds, or a time span value such | |
773 | as "5min 20s". Pass <literal>0</literal> to disable the | |
774 | timeout logic. Defaults to | |
775 | <varname>DefaultTimeoutStartSec=</varname> from the manager | |
776 | configuration file (see | |
777 | <citerefentry><refentrytitle>systemd-system.conf</refentrytitle><manvolnum>5</manvolnum></citerefentry>). | |
778 | </para></listitem> | |
779 | </varlistentry> | |
780 | ||
781 | <varlistentry> | |
782 | <term><varname>Service=</varname></term> | |
783 | <listitem><para>Specifies the service unit name to activate on | |
784 | incoming traffic. This setting is only allowed for sockets | |
785 | with <varname>Accept=no</varname>. It defaults to the service | |
786 | that bears the same name as the socket (with the suffix | |
787 | replaced). In most cases, it should not be necessary to use | |
db2df898 MP |
788 | this option. Note that setting this parameter might result in |
789 | additional dependencies to be added to the unit (see | |
790 | above).</para></listitem> | |
e735f4d4 MP |
791 | </varlistentry> |
792 | ||
793 | <varlistentry> | |
794 | <term><varname>RemoveOnStop=</varname></term> | |
795 | <listitem><para>Takes a boolean argument. If enabled, any file | |
796 | nodes created by this socket unit are removed when it is | |
797 | stopped. This applies to AF_UNIX sockets in the file system, | |
798 | POSIX message queues, FIFOs, as well as any symlinks to them | |
799 | configured with <varname>Symlinks=</varname>. Normally, it | |
800 | should not be necessary to use this option, and is not | |
801 | recommended as services might continue to run after the socket | |
802 | unit has been terminated and it should still be possible to | |
803 | communicate with them via their file system node. Defaults to | |
804 | off.</para></listitem> | |
805 | </varlistentry> | |
806 | ||
807 | <varlistentry> | |
808 | <term><varname>Symlinks=</varname></term> | |
f5e65279 MB |
809 | <listitem><para>Takes a list of file system paths. The specified paths will be created as symlinks to the |
810 | <constant>AF_UNIX</constant> socket path or FIFO path of this socket unit. If this setting is used, only one | |
811 | <constant>AF_UNIX</constant> socket in the file system or one FIFO may be configured for the socket unit. Use | |
812 | this option to manage one or more symlinked alias names for a socket, binding their lifecycle together. Note | |
813 | that if creation of a symlink fails this is not considered fatal for the socket unit, and the socket unit may | |
814 | still start. If an empty string is assigned, the list of paths is reset. Defaults to an empty | |
e735f4d4 MP |
815 | list.</para></listitem> |
816 | </varlistentry> | |
817 | ||
6300502b MP |
818 | <varlistentry> |
819 | <term><varname>FileDescriptorName=</varname></term> | |
820 | <listitem><para>Assigns a name to all file descriptors this | |
821 | socket unit encapsulates. This is useful to help activated | |
db2df898 | 822 | services identify specific file descriptors, if multiple fds |
6300502b MP |
823 | are passed. Services may use the |
824 | <citerefentry><refentrytitle>sd_listen_fds_with_names</refentrytitle><manvolnum>3</manvolnum></citerefentry> | |
825 | call to acquire the names configured for the received file | |
826 | descriptors. Names may contain any ASCII character, but must | |
db2df898 | 827 | exclude control characters and <literal>:</literal>, and must |
6300502b | 828 | be at most 255 characters in length. If this setting is not |
db2df898 | 829 | used, the file descriptor name defaults to the name of the |
6300502b MP |
830 | socket unit, including its <filename>.socket</filename> |
831 | suffix.</para></listitem> | |
832 | </varlistentry> | |
833 | ||
aa27b158 MP |
834 | <varlistentry> |
835 | <term><varname>TriggerLimitIntervalSec=</varname></term> | |
836 | <term><varname>TriggerLimitBurst=</varname></term> | |
837 | ||
838 | <listitem><para>Configures a limit on how often this socket unit my be activated within a specific time | |
839 | interval. The <varname>TriggerLimitIntervalSec=</varname> may be used to configure the length of the time | |
840 | interval in the usual time units <literal>us</literal>, <literal>ms</literal>, <literal>s</literal>, | |
841 | <literal>min</literal>, <literal>h</literal>, … and defaults to 2s (See | |
842 | <citerefentry><refentrytitle>systemd.time</refentrytitle><manvolnum>7</manvolnum></citerefentry> for details on | |
843 | the various time units understood). The <varname>TriggerLimitBurst=</varname> setting takes a positive integer | |
844 | value and specifies the number of permitted activations per time interval, and defaults to 200 for | |
845 | <varname>Accept=yes</varname> sockets (thus by default permitting 200 activations per 2s), and 20 otherwise (20 | |
846 | activations per 2s). Set either to 0 to disable any form of trigger rate limiting. If the limit is hit, the | |
847 | socket unit is placed into a failure mode, and will not be connectible anymore until restarted. Note that this | |
848 | limit is enforced before the service activation is enqueued.</para></listitem> | |
849 | </varlistentry> | |
850 | ||
e735f4d4 MP |
851 | </variablelist> |
852 | ||
853 | <para>Check | |
854 | <citerefentry><refentrytitle>systemd.exec</refentrytitle><manvolnum>5</manvolnum></citerefentry> | |
855 | and | |
856 | <citerefentry><refentrytitle>systemd.kill</refentrytitle><manvolnum>5</manvolnum></citerefentry> | |
857 | for more settings.</para> | |
858 | ||
859 | </refsect1> | |
860 | ||
861 | <refsect1> | |
862 | <title>See Also</title> | |
863 | <para> | |
864 | <citerefentry><refentrytitle>systemd</refentrytitle><manvolnum>1</manvolnum></citerefentry>, | |
865 | <citerefentry><refentrytitle>systemctl</refentrytitle><manvolnum>1</manvolnum></citerefentry>, | |
866 | <citerefentry><refentrytitle>systemd.unit</refentrytitle><manvolnum>5</manvolnum></citerefentry>, | |
867 | <citerefentry><refentrytitle>systemd.exec</refentrytitle><manvolnum>5</manvolnum></citerefentry>, | |
868 | <citerefentry><refentrytitle>systemd.kill</refentrytitle><manvolnum>5</manvolnum></citerefentry>, | |
869 | <citerefentry><refentrytitle>systemd.resource-control</refentrytitle><manvolnum>5</manvolnum></citerefentry>, | |
870 | <citerefentry><refentrytitle>systemd.service</refentrytitle><manvolnum>5</manvolnum></citerefentry>, | |
6300502b MP |
871 | <citerefentry><refentrytitle>systemd.directives</refentrytitle><manvolnum>7</manvolnum></citerefentry>, |
872 | <citerefentry><refentrytitle>sd_listen_fds</refentrytitle><manvolnum>3</manvolnum></citerefentry>, | |
873 | <citerefentry><refentrytitle>sd_listen_fds_with_names</refentrytitle><manvolnum>3</manvolnum></citerefentry> | |
e735f4d4 | 874 | </para> |
e735f4d4 MP |
875 | <para> |
876 | For more extensive descriptions see the "systemd for Developers" series: | |
877 | <ulink url="http://0pointer.de/blog/projects/socket-activation.html">Socket Activation</ulink>, | |
878 | <ulink url="http://0pointer.de/blog/projects/socket-activation2.html">Socket Activation, part II</ulink>, | |
879 | <ulink url="http://0pointer.de/blog/projects/inetd.html">Converting inetd Services</ulink>, | |
880 | <ulink url="http://0pointer.de/blog/projects/socket-activated-containers.html">Socket Activated Internet Services and OS Containers</ulink>. | |
881 | </para> | |
882 | </refsect1> | |
663996b3 MS |
883 | |
884 | </refentry> |