Debian Codegraph @

upstream/mageia-cauldron/man5/systemd.socket.5

raw
1 '\" t
2 .TH "SYSTEMD\&.SOCKET" "5" "" "systemd 258.7" "systemd.socket"
3 .\" -----------------------------------------------------------------
4 .\" * Define some portability stuff
5 .\" -----------------------------------------------------------------
6 .\" ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
7 .\" http://bugs.debian.org/507673
8 .\" http://lists.gnu.org/archive/html/groff/2009-02/msg00013.html
9 .\" ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
10 .ie \n(.g .ds Aq \(aq
11 .el .ds Aq '
12 .\" -----------------------------------------------------------------
13 .\" * set default formatting
14 .\" -----------------------------------------------------------------
15 .\" disable hyphenation
16 .nh
17 .\" disable justification (adjust text to left margin only)
18 .ad l
19 .\" -----------------------------------------------------------------
20 .\" * MAIN CONTENT STARTS HERE *
21 .\" -----------------------------------------------------------------
22 .SH "NAME"
23 systemd.socket \- Socket unit configuration
24 .SH "SYNOPSIS"
25 .PP
26 \fIsocket\fR\&.socket
27 .SH "DESCRIPTION"
28 .PP
29 A unit configuration file whose name ends in
30 "\&.socket"
31 encodes information about an IPC or network socket or a file system FIFO controlled and supervised by systemd, for socket\-based activation\&.
32 .PP
33 This man page lists the configuration options specific to this unit type\&. See
34 \fBsystemd.unit\fR(5)
35 for the common options of all unit configuration files\&. The common configuration items are configured in the generic [Unit] and [Install] sections\&. The socket specific configuration options are configured in the [Socket] section\&.
36 .PP
37 Additional options are listed in
38 \fBsystemd.exec\fR(5), which define the execution environment the
39 \fBExecStartPre=\fR,
40 \fBExecStartPost=\fR,
41 \fBExecStopPre=\fR
42 and
43 \fBExecStopPost=\fR
44 commands are executed in, and in
45 \fBsystemd.kill\fR(5), which define the way the processes are terminated, and in
46 \fBsystemd.resource-control\fR(5), which configure resource control settings for the processes of the socket\&.
47 .PP
48 For each socket unit, a matching service unit must exist, describing the service to start on incoming traffic on the socket (see
49 \fBsystemd.service\fR(5)
50 for more information about \&.service units)\&. The name of the \&.service unit is by default the same as the name of the \&.socket unit, but can be altered with the
51 \fBService=\fR
52 option described below\&. Depending on the setting of the
53 \fBAccept=\fR
54 option described below, this \&.service unit must either be named like the \&.socket unit, but with the suffix replaced, unless overridden with
55 \fBService=\fR; or it must be a template unit named the same way\&. Example: a socket file
56 foo\&.socket
57 needs a matching service
58 foo\&.service
59 if
60 \fBAccept=no\fR
61 is set\&. If
62 \fBAccept=yes\fR
63 is set, a service template
64 foo@\&.service
65 must exist from which services are instantiated for each incoming connection\&.
66 .PP
67 No implicit
68 \fIWantedBy=\fR
69 or
70 \fIRequiredBy=\fR
71 dependency from the socket to the service is added\&. This means that the service may be started without the socket, in which case it must be able to open sockets by itself\&. To prevent this, an explicit
72 \fIRequires=\fR
73 dependency may be added\&.
74 .PP
75 Socket units may be used to implement on\-demand starting of services, as well as parallelized starting of services\&. See the blog stories linked at the end for an introduction\&.
76 .PP
77 Note that the daemon software configured for socket activation with socket units needs to be able to accept sockets from systemd, either via systemd\*(Aqs native socket passing interface (see
78 \fBsd_listen_fds\fR(3)
79 for details about the precise protocol used and the order in which the file descriptors are passed) or via traditional
80 \fBinetd\fR(8)\-style socket passing (i\&.e\&. sockets passed in via standard input and output, using
81 \fIStandardInput=socket\fR
82 in the service file)\&.
83 .PP
84 All network sockets allocated through
85 \&.socket
86 units are allocated in the host\*(Aqs network namespace (see
87 \fBnetwork_namespaces\fR(7))\&. This does not mean however that the service activated by a configured socket unit has to be part of the host\*(Aqs network namespace as well\&. It is supported and even good practice to run services in their own network namespace (for example through
88 \fIPrivateNetwork=\fR, see
89 \fBsystemd.exec\fR(5)), receiving only the sockets configured through socket\-activation from the host\*(Aqs namespace\&. In such a set\-up communication within the host\*(Aqs network namespace is only permitted through the activation sockets passed in while all sockets allocated from the service code itself will be associated with the service\*(Aqs own namespace, and thus possibly subject to a restrictive configuration\&.
90 .SH "AUTOMATIC DEPENDENCIES"
91 .SS "Implicit Dependencies"
92 .PP
93 The following dependencies are implicitly added:
94 .sp
95 .RS 4
96 .ie n \{\
97 \h'-04'\(bu\h'+03'\c
98 .\}
99 .el \{\
100 .sp -1
101 .IP \(bu 2.3
102 .\}
103 Socket units automatically gain a
104 \fIBefore=\fR
105 dependency on the service units they activate\&.
106 .RE
107 .sp
108 .RS 4
109 .ie n \{\
110 \h'-04'\(bu\h'+03'\c
111 .\}
112 .el \{\
113 .sp -1
114 .IP \(bu 2.3
115 .\}
116 Socket units referring to file system paths (such as
117 \fBAF_UNIX\fR
118 sockets or FIFOs) implicitly gain
119 \fIRequires=\fR
120 and
121 \fIAfter=\fR
122 dependencies on all mount units necessary to access those paths\&.
123 .RE
124 .sp
125 .RS 4
126 .ie n \{\
127 \h'-04'\(bu\h'+03'\c
128 .\}
129 .el \{\
130 .sp -1
131 .IP \(bu 2.3
132 .\}
133 Socket units using the
134 \fIBindToDevice=\fR
135 setting automatically gain a
136 \fIBindsTo=\fR
137 and
138 \fIAfter=\fR
139 dependency on the device unit encapsulating the specified network interface\&.
140 .RE
141 .PP
142 Additional implicit dependencies may be added as result of execution and resource control parameters as documented in
143 \fBsystemd.exec\fR(5)
144 and
145 \fBsystemd.resource-control\fR(5)\&.
146 .SS "Default Dependencies"
147 .PP
148 The following dependencies are added unless
149 \fIDefaultDependencies=no\fR
150 is set:
151 .sp
152 .RS 4
153 .ie n \{\
154 \h'-04'\(bu\h'+03'\c
155 .\}
156 .el \{\
157 .sp -1
158 .IP \(bu 2.3
159 .\}
160 Socket units automatically gain a
161 \fIBefore=\fR
162 dependency on
163 sockets\&.target\&.
164 .RE
165 .sp
166 .RS 4
167 .ie n \{\
168 \h'-04'\(bu\h'+03'\c
169 .\}
170 .el \{\
171 .sp -1
172 .IP \(bu 2.3
173 .\}
174 Socket units automatically gain a pair of
175 \fIAfter=\fR
176 and
177 \fIRequires=\fR
178 dependency on
179 sysinit\&.target, and a pair of
180 \fIBefore=\fR
181 and
182 \fIConflicts=\fR
183 dependencies on
184 shutdown\&.target\&. These dependencies ensure that the socket unit is started before normal services at boot, and is stopped on shutdown\&. Only sockets involved with early boot or late system shutdown should disable
185 \fIDefaultDependencies=\fR
186 option\&.
187 .RE
188 .SH "OPTIONS"
189 .PP
190 Socket unit files may include [Unit] and [Install] sections, which are described in
191 \fBsystemd.unit\fR(5)\&.
192 .PP
193 Socket unit files must include a [Socket] section, which carries information about the socket or FIFO it supervises\&. A number of options that may be used in this section are shared with other unit types\&. These options are documented in
194 \fBsystemd.exec\fR(5),
195 \fBsystemd.kill\fR(5)
196 and
197 \fBsystemd.resource-control\fR(5)\&. The options specific to the [Socket] section of socket units are the following:
198 .PP
199 \fIListenStream=\fR, \fIListenDatagram=\fR, \fIListenSequentialPacket=\fR
200 .RS 4
201 Specifies an address to listen on for a stream (\fBSOCK_STREAM\fR), datagram (\fBSOCK_DGRAM\fR), or sequential packet (\fBSOCK_SEQPACKET\fR) socket, respectively\&. The address can be written in various formats:
202 .sp
203 If the address starts with a slash ("/"), it is read as file system socket in the
204 \fBAF_UNIX\fR
205 socket family\&.
206 .sp
207 If the address starts with an at symbol ("@"), it is read as abstract namespace socket in the
208 \fBAF_UNIX\fR
209 family\&. The
210 "@"
211 is replaced with a
212 \fBNUL\fR
213 character before binding\&. For details, see
214 \fBunix\fR(7)\&.
215 .sp
216 If the address string is a single number, it is read as port number to listen on via IPv6\&. Depending on the value of
217 \fIBindIPv6Only=\fR
218 (see below) this might result in the service being available via both IPv6 and IPv4 (default) or just via IPv6\&.
219 .sp
220 If the address string is a string in the format
221 "\fIv\&.w\&.x\&.y\fR:\fIz\fR", it is interpreted as IPv4 address
222 \fIv\&.w\&.x\&.y\fR
223 and port
224 \fIz\fR\&.
225 .sp
226 If the address string is a string in the format
227 "[\fIx\fR]:\fIy\fR", it is interpreted as IPv6 address
228 \fIx\fR
229 and port
230 \fIy\fR\&. An optional interface scope (interface name or number) may be specified after a
231 "%"
232 symbol:
233 "[\fIx\fR]:\fIy\fR%\fIdev\fR"\&. Interface scopes are only useful with link\-local addresses, because the kernel ignores them in other cases\&. Note that if an address is specified as IPv6, it might still make the service available via IPv4 too, depending on the
234 \fIBindIPv6Only=\fR
235 setting (see below)\&.
236 .sp
237 If the address string is a string in the format
238 "vsock:\fIx\fR:\fIy\fR", it is read as CID
239 \fIx\fR
240 on a port
241 \fIy\fR
242 address in the
243 \fBAF_VSOCK\fR
244 family\&. The CID is a unique 32\-bit integer identifier in
245 \fBAF_VSOCK\fR
246 analogous to an IP address\&. Specifying the CID is optional, and may be set to the empty string\&.
247 "vsock"
248 may be replaced with
249 "vsock\-stream",
250 "vsock\-dgram"
251 or
252 "vsock\-seqpacket"
253 to force usage of the corresponding socket type\&.
254 .sp
255 Note that
256 \fBSOCK_SEQPACKET\fR
257 (i\&.e\&.
258 \fIListenSequentialPacket=\fR) is only available for
259 \fBAF_UNIX\fR
260 sockets\&.
261 \fBSOCK_STREAM\fR
262 (i\&.e\&.
263 \fIListenStream=\fR) when used for IP sockets refers to TCP sockets,
264 \fBSOCK_DGRAM\fR
265 (i\&.e\&.
266 \fIListenDatagram=\fR) to UDP\&.
267 .sp
268 These options may be specified more than once, in which case incoming traffic on any of the sockets will trigger service activation, and all listed sockets will be passed to the service, regardless of whether there is incoming traffic on them or not\&. If the empty string is assigned to any of these options, the list of addresses to listen on is reset, all prior uses of any of these options will have no effect\&.
269 .sp
270 It is also possible to have more than one socket unit for the same service when using
271 \fIService=\fR, and the service will receive all the sockets configured in all the socket units\&. Sockets configured in one unit are passed in the order of configuration, but no ordering between socket units is specified\&.
272 .sp
273 If an IP address is used here, it is often desirable to listen on it before the interface it is configured on is up and running, and even regardless of whether it will be up and running at any point\&. To deal with this, it is recommended to set the
274 \fIFreeBind=\fR
275 option described below\&.
276 .RE
277 .PP
278 \fIListenFIFO=\fR
279 .RS 4
280 Specifies a file system FIFO (see
281 \fBfifo\fR(7)
282 for details) to listen on\&. This expects an absolute file system path as argument\&. Behavior otherwise is very similar to the
283 \fIListenDatagram=\fR
284 directive above\&.
285 .RE
286 .PP
287 \fIListenSpecial=\fR
288 .RS 4
289 Specifies a special file in the file system to listen on\&. This expects an absolute file system path as argument\&. Behavior otherwise is very similar to the
290 \fIListenFIFO=\fR
291 directive above\&. Use this to open character device nodes as well as special files in
292 /proc/
293 and
294 /sys/\&.
295 .RE
296 .PP
297 \fIListenNetlink=\fR
298 .RS 4
299 Specifies a Netlink family to create a socket for to listen on\&. This expects a short string referring to the
300 \fBAF_NETLINK\fR
301 family name (such as
302 \fIaudit\fR
303 or
304 \fIkobject\-uevent\fR) as argument, optionally suffixed by a whitespace followed by a multicast group integer\&. Behavior otherwise is very similar to the
305 \fIListenDatagram=\fR
306 directive above\&.
307 .RE
308 .PP
309 \fIListenMessageQueue=\fR
310 .RS 4
311 Specifies a POSIX message queue name to listen on (see
312 \fBmq_overview\fR(7)
313 for details)\&. This expects a valid message queue name (i\&.e\&. beginning with
314 "/")\&. Behavior otherwise is very similar to the
315 \fIListenFIFO=\fR
316 directive above\&. On Linux message queue descriptors are actually file descriptors and can be inherited between processes\&.
317 .RE
318 .PP
319 \fIListenUSBFunction=\fR
320 .RS 4
321 Specifies a
322 \m[blue]\fBUSB FunctionFS\fR\m[]\&\s-2\u[1]\d\s+2
323 endpoints location to listen on, for implementation of USB gadget functions\&. This expects an absolute file system path of a FunctionFS mount point as the argument\&. Behavior otherwise is very similar to the
324 \fIListenFIFO=\fR
325 directive above\&. Use this to open the FunctionFS endpoint
326 ep0\&. When using this option, the activated service has to have the
327 \fIUSBFunctionDescriptors=\fR
328 and
329 \fIUSBFunctionStrings=\fR
330 options set\&.
331 .sp
332 Added in version 227\&.
333 .RE
334 .PP
335 \fISocketProtocol=\fR
336 .RS 4
337 Takes one of
338 \fBudplite\fR,
339 \fBsctp\fR
340 or
341 \fBmptcp\fR\&. The socket will use the UDP\-Lite (\fBIPPROTO_UDPLITE\fR), SCTP (\fBIPPROTO_SCTP\fR) or MPTCP (\fBIPPROTO_MPTCP\fR) protocol, respectively\&.
342 .sp
343 Added in version 229\&.
344 .RE
345 .PP
346 \fIBindIPv6Only=\fR
347 .RS 4
348 Takes one of
349 \fBdefault\fR,
350 \fBboth\fR
351 or
352 \fBipv6\-only\fR\&. Controls the IPV6_V6ONLY socket option (see
353 \fBipv6\fR(7)
354 for details)\&. If
355 \fBboth\fR, IPv6 sockets bound will be accessible via both IPv4 and IPv6\&. If
356 \fBipv6\-only\fR, they will be accessible via IPv6 only\&. If
357 \fBdefault\fR
358 (which is the default, surprise!), the system wide default setting is used, as controlled by
359 /proc/sys/net/ipv6/bindv6only, which in turn defaults to the equivalent of
360 \fBboth\fR\&.
361 .RE
362 .PP
363 \fIBacklog=\fR
364 .RS 4
365 Takes an unsigned 32\-bit integer argument\&. Specifies the number of connections to queue that have not been accepted yet\&. This setting matters only for stream and sequential packet sockets\&. See
366 \fBlisten\fR(2)
367 for details\&. Defaults to 4294967295\&. Note that this value is silently capped by the
368 "net\&.core\&.somaxconn"
369 sysctl, which typically defaults to 4096, so typically the sysctl is the setting that actually matters\&.
370 .RE
371 .PP
372 \fIBindToDevice=\fR
373 .RS 4
374 Specifies a network interface name to bind this socket to\&. If set, traffic will only be accepted from the specified network interfaces\&. This controls the
375 \fBSO_BINDTODEVICE\fR
376 socket option (see
377 \fBsocket\fR(7)
378 for details)\&. If this option is used, an implicit dependency from this socket unit on the network interface device unit is created (see
379 \fBsystemd.device\fR(5))\&. Note that setting this parameter might result in additional dependencies to be added to the unit (see above)\&.
380 .RE
381 .PP
382 \fISocketUser=\fR, \fISocketGroup=\fR
383 .RS 4
384 Takes a UNIX user/group name\&. When specified, all
385 \fBAF_UNIX\fR
386 sockets, FIFO nodes, and message queues are owned by the specified user and group\&. If unset (the default), the nodes are owned by the root user/group (if run in system context) or the invoking user/group (if run in user context)\&. If only a user is specified but no group, then the group is derived from the user\*(Aqs default group\&.
387 .sp
388 Added in version 214\&.
389 .RE
390 .PP
391 \fISocketMode=\fR
392 .RS 4
393 If listening on a file system socket, FIFO, or message queue, this option specifies the file system access mode used when creating the file node\&. Takes an access mode in octal notation\&. Defaults to 0666\&.
394 .RE
395 .PP
396 \fIDirectoryMode=\fR
397 .RS 4
398 If listening on a file system socket or FIFO, the parent directories are automatically created if needed\&. This option specifies the file system access mode used when creating these directories\&. Takes an access mode in octal notation\&. Defaults to 0755\&.
399 .RE
400 .PP
401 \fIAccept=\fR
402 .RS 4
403 Takes a boolean argument\&. If yes, a service instance is spawned for each incoming connection and only the connection socket is passed to it\&. If no, all listening sockets themselves are passed to the started service unit, and only one service unit is spawned for all connections (also see above)\&. This value is ignored for datagram sockets and FIFOs where a single service unit unconditionally handles all incoming traffic\&. Defaults to
404 \fBno\fR\&.
405 .sp
406 Typically, for performance sensitive services, a choice of
407 \fBAccept=no\fR
408 is preferable, since that way only the first connection will have to pay the activation resource cost\&. On the other hand, for sporadically used services
409 \fBAccept=yes\fR
410 can be preferable as it simplifies the implementation (as the service program code only has to process a single connection instead of handling multiple) and enables stronger security (since the various sandboxing options can be used to isolate parallel connections from each other, as each is serviced by a separate service instance and process)\&.
411 .sp
412 A service listening on an
413 \fBAF_UNIX\fR
414 socket may, but does not need to, call
415 \fBclose\fR(2)
416 on the received socket before exiting\&. However, it must not unlink the socket from a file system\&. It should not invoke
417 \fBshutdown\fR(2)
418 on sockets it got with
419 \fIAccept=no\fR, but it may do so for sockets it got with
420 \fIAccept=yes\fR
421 set\&.
422 .sp
423 Setting
424 \fIAccept=yes\fR
425 is in particular useful for allowing daemons designed for usage with
426 \fBinetd\fR(8)
427 to work unmodified with systemd socket activation\&.
428 .sp
429 Note that depending on this setting the services activated by units of this type are either regular services (in case of
430 \fIAccept=\fR\fBno\fR) or instances of templated services (in case of
431 \fIAccept=\fR\fByes\fR)\&. See the Description section above for a more detailed discussion of the naming rules of triggered services\&.
432 .sp
433 For IPv4 and IPv6 connections, the
434 \fI$REMOTE_ADDR\fR
435 environment variable will contain the remote IP address, and
436 \fI$REMOTE_PORT\fR
437 will contain the remote port number\&. These two variables correspond to those defined by the CGI interface for web services (see
438 \m[blue]\fBRFC 3875\fR\m[]\&\s-2\u[2]\d\s+2)\&.
439 .sp
440 For
441 \fBAF_UNIX\fR
442 socket connections, the
443 \fI$REMOTE_ADDR\fR
444 environment variable will contain either the remote socket\*(Aqs file system path starting with a slash ("/") or its address in the abstract namespace starting with an at symbol ("@")\&. If the socket is unnamed,
445 \fI$REMOTE_ADDR\fR
446 will not be set\&.
447 .sp
448 If
449 \fIAccept=yes\fR
450 is used, the activated service process will have set the
451 \fI$SO_COOKIE\fR
452 environment variable to the Linux socket cookie, formatted as decimal integer\&. The socket cookie can otherwise be acquired via
453 \fBgetsockopt\fR(7)\&.
454 .sp
455 It is recommended to set
456 \fICollectMode=inactive\-or\-failed\fR
457 for service instances activated via
458 \fIAccept=yes\fR, to ensure that failed connection services are cleaned up and released from memory, and do not accumulate\&.
459 .RE
460 .PP
461 \fIWritable=\fR
462 .RS 4
463 Takes a boolean argument\&. May only be used in conjunction with
464 \fIListenSpecial=\fR\&. If true, the specified special file is opened in read\-write mode, if false, in read\-only mode\&. Defaults to false\&.
465 .sp
466 Added in version 227\&.
467 .RE
468 .PP
469 \fIFlushPending=\fR
470 .RS 4
471 Takes a boolean argument\&. May only be used when
472 \fBAccept=no\fR\&. If yes, the socket\*(Aqs buffers are cleared after the triggered service exited\&. This causes any pending data to be flushed and any pending incoming connections to be rejected\&. If no, the socket\*(Aqs buffers will not be cleared, permitting the service to handle any pending connections after restart, which is the usually expected behaviour\&. Defaults to
473 \fBno\fR\&.
474 .sp
475 Added in version 247\&.
476 .RE
477 .PP
478 \fIMaxConnections=\fR
479 .RS 4
480 The maximum number of connections to simultaneously run services instances for, when
481 \fBAccept=yes\fR
482 is set\&. If more concurrent connections are coming in, they will be refused until at least one existing connection is terminated\&. This setting has no effect on sockets configured with
483 \fBAccept=no\fR
484 or datagram sockets\&. Defaults to 64\&.
485 .RE
486 .PP
487 \fIMaxConnectionsPerSource=\fR
488 .RS 4
489 The maximum number of connections for a service per source IP address (in case of IPv4/IPv6), per source CID (in case of
490 \fBAF_VSOCK\fR), or source UID (in case of
491 \fBAF_UNIX\fR)\&. This is very similar to the
492 \fIMaxConnections=\fR
493 directive above\&. Defaults to 0, i\&.e\&. disabled\&.
494 .sp
495 Added in version 232\&.
496 .RE
497 .PP
498 \fIKeepAlive=\fR
499 .RS 4
500 Takes a boolean argument\&. If true, the TCP/IP stack will send a keep alive message after 2h (depending on the configuration of
501 /proc/sys/net/ipv4/tcp_keepalive_time) for all TCP streams accepted on this socket\&. This controls the
502 \fBSO_KEEPALIVE\fR
503 socket option (see
504 \fBsocket\fR(7)
505 and the
506 \m[blue]\fBTCP Keepalive HOWTO\fR\m[]\&\s-2\u[3]\d\s+2
507 for details\&.) Defaults to
508 \fBfalse\fR\&.
509 .RE
510 .PP
511 \fIKeepAliveTimeSec=\fR
512 .RS 4
513 Takes time (in seconds) as argument\&. The connection needs to remain idle before TCP starts sending keepalive probes\&. This controls the TCP_KEEPIDLE socket option (see
514 \fBsocket\fR(7)
515 and the
516 \m[blue]\fBTCP Keepalive HOWTO\fR\m[]\&\s-2\u[3]\d\s+2
517 for details\&.) Default value is 7200 seconds (2 hours)\&.
518 .sp
519 Added in version 216\&.
520 .RE
521 .PP
522 \fIKeepAliveIntervalSec=\fR
523 .RS 4
524 Takes time (in seconds) as argument between individual keepalive probes, if the socket option
525 \fBSO_KEEPALIVE\fR
526 has been set on this socket\&. This controls the
527 \fBTCP_KEEPINTVL\fR
528 socket option (see
529 \fBsocket\fR(7)
530 and the
531 \m[blue]\fBTCP Keepalive HOWTO\fR\m[]\&\s-2\u[3]\d\s+2
532 for details\&.) Default value is 75 seconds\&.
533 .sp
534 Added in version 216\&.
535 .RE
536 .PP
537 \fIKeepAliveProbes=\fR
538 .RS 4
539 Takes an integer as argument\&. It is the number of unacknowledged probes to send before considering the connection dead and notifying the application layer\&. This controls the TCP_KEEPCNT socket option (see
540 \fBsocket\fR(7)
541 and the
542 \m[blue]\fBTCP Keepalive HOWTO\fR\m[]\&\s-2\u[3]\d\s+2
543 for details\&.) Default value is 9\&.
544 .sp
545 Added in version 216\&.
546 .RE
547 .PP
548 \fINoDelay=\fR
549 .RS 4
550 Takes a boolean argument\&. TCP Nagle\*(Aqs algorithm works by combining a number of small outgoing messages, and sending them all at once\&. This controls the TCP_NODELAY socket option (see
551 \fBtcp\fR(7))\&. Defaults to
552 \fBfalse\fR\&.
553 .sp
554 Added in version 216\&.
555 .RE
556 .PP
557 \fIPriority=\fR
558 .RS 4
559 Takes an integer argument controlling the priority for all traffic sent from this socket\&. This controls the
560 \fBSO_PRIORITY\fR
561 socket option (see
562 \fBsocket\fR(7)
563 for details\&.)\&.
564 .RE
565 .PP
566 \fIDeferAcceptSec=\fR
567 .RS 4
568 Takes time (in seconds) as argument\&. If set, the listening process will be awakened only when data arrives on the socket, and not immediately when connection is established\&. When this option is set, the
569 \fBTCP_DEFER_ACCEPT\fR
570 socket option will be used (see
571 \fBtcp\fR(7)), and the kernel will ignore initial ACK packets without any data\&. The argument specifies the approximate amount of time the kernel should wait for incoming data before falling back to the normal behavior of honoring empty ACK packets\&. This option is beneficial for protocols where the client sends the data first (e\&.g\&. HTTP, in contrast to SMTP), because the server process will not be woken up unnecessarily before it can take any action\&.
572 .sp
573 If the client also uses the
574 \fBTCP_DEFER_ACCEPT\fR
575 option, the latency of the initial connection may be reduced, because the kernel will send data in the final packet establishing the connection (the third packet in the "three\-way handshake")\&.
576 .sp
577 Disabled by default\&.
578 .sp
579 Added in version 216\&.
580 .RE
581 .PP
582 \fIReceiveBuffer=\fR, \fISendBuffer=\fR
583 .RS 4
584 Takes an integer argument controlling the receive or send buffer sizes of this socket, respectively\&. This controls the
585 \fBSO_RCVBUF\fR
586 and
587 \fBSO_SNDBUF\fR
588 socket options (see
589 \fBsocket\fR(7)
590 for details\&.)\&. The usual suffixes K, M, G are supported and are understood to the base of 1024\&.
591 .RE
592 .PP
593 \fIIPTOS=\fR
594 .RS 4
595 Takes an integer argument controlling the IP Type\-Of\-Service field for packets generated from this socket\&. This controls the
596 \fBIP_TOS\fR
597 socket option (see
598 \fBip\fR(7)
599 for details\&.)\&. Either a numeric string or one of
600 \fBlow\-delay\fR,
601 \fBthroughput\fR,
602 \fBreliability\fR
603 or
604 \fBlow\-cost\fR
605 may be specified\&.
606 .RE
607 .PP
608 \fIIPTTL=\fR
609 .RS 4
610 Takes an integer argument controlling the IPv4 Time\-To\-Live/IPv6 Hop\-Count field for packets generated from this socket\&. This sets the
611 \fBIP_TTL\fR/\fBIPV6_UNICAST_HOPS\fR
612 socket options (see
613 \fBip\fR(7)
614 and
615 \fBipv6\fR(7)
616 for details\&.)
617 .RE
618 .PP
619 \fIMark=\fR
620 .RS 4
621 Takes an integer value\&. Controls the firewall mark of packets generated by this socket\&. This can be used in the firewall logic to filter packets from this socket\&. This sets the
622 \fBSO_MARK\fR
623 socket option\&. See
624 \fBiptables\fR(8)
625 for details\&.
626 .RE
627 .PP
628 \fIReusePort=\fR
629 .RS 4
630 Takes a boolean value\&. If true, allows multiple
631 \fBbind\fR(2)s to this TCP or UDP port\&. This controls the
632 \fBSO_REUSEPORT\fR
633 socket option\&. See
634 \fBsocket\fR(7)
635 for details\&.
636 .sp
637 Added in version 206\&.
638 .RE
639 .PP
640 \fISmackLabel=\fR, \fISmackLabelIPIn=\fR, \fISmackLabelIPOut=\fR
641 .RS 4
642 Takes a string value\&. Controls the extended attributes
643 "security\&.SMACK64",
644 "security\&.SMACK64IPIN"
645 and
646 "security\&.SMACK64IPOUT", respectively, i\&.e\&. the security label of the FIFO, or the security label for the incoming or outgoing connections of the socket, respectively\&. See
647 \m[blue]\fBSmack\fR\m[]\&\s-2\u[4]\d\s+2
648 for details\&.
649 .sp
650 Added in version 196\&.
651 .RE
652 .PP
653 \fISELinuxContextFromNet=\fR
654 .RS 4
655 Takes a boolean argument\&. When true, systemd will attempt to figure out the SELinux label used for the instantiated service from the information handed by the peer over the network\&. Note that only the security level is used from the information provided by the peer\&. Other parts of the resulting SELinux context originate from either the target binary that is effectively triggered by socket unit or from the value of the
656 \fISELinuxContext=\fR
657 option\&. This configuration option applies only when activated service is passed in single socket file descriptor, i\&.e\&. service instances that have standard input connected to a socket or services triggered by exactly one socket unit\&. Also note that this option is useful only when MLS/MCS SELinux policy is deployed\&. Defaults to
658 "false"\&.
659 .sp
660 Added in version 217\&.
661 .RE
662 .PP
663 \fIPipeSize=\fR
664 .RS 4
665 Takes a size in bytes\&. Controls the pipe buffer size of FIFOs configured in this socket unit\&. See
666 \fBfcntl\fR(2)
667 for details\&. The usual suffixes K, M, G are supported and are understood to the base of 1024\&.
668 .RE
669 .PP
670 \fIMessageQueueMaxMessages=\fR, \fIMessageQueueMessageSize=\fR
671 .RS 4
672 These two settings take integer values and control the mq_maxmsg field or the mq_msgsize field, respectively, when creating the message queue\&. Note that either none or both of these variables need to be set\&. See
673 \fBmq_setattr\fR(3)
674 for details\&.
675 .RE
676 .PP
677 \fIFreeBind=\fR
678 .RS 4
679 Takes a boolean value\&. Controls whether the socket can be bound to non\-local IP addresses\&. This is useful to configure sockets listening on specific IP addresses before those IP addresses are successfully configured on a network interface\&. This sets the
680 \fBIP_FREEBIND\fR/\fBIPV6_FREEBIND\fR
681 socket option\&. For robustness reasons it is recommended to use this option whenever you bind a socket to a specific IP address\&. Defaults to
682 \fBfalse\fR\&.
683 .RE
684 .PP
685 \fITransparent=\fR
686 .RS 4
687 Takes a boolean value\&. Controls the
688 \fBIP_TRANSPARENT\fR/\fBIPV6_TRANSPARENT\fR
689 socket option\&. Defaults to
690 \fBfalse\fR\&.
691 .RE
692 .PP
693 \fIBroadcast=\fR
694 .RS 4
695 Takes a boolean value\&. This controls the
696 \fBSO_BROADCAST\fR
697 socket option, which allows broadcast datagrams to be sent from this socket\&. Defaults to
698 \fBfalse\fR\&.
699 .RE
700 .PP
701 \fIPassCredentials=\fR
702 .RS 4
703 Takes a boolean value\&. This controls the
704 \fBSO_PASSCRED\fR
705 socket option, which allows
706 \fBAF_UNIX\fR
707 sockets to receive the credentials of the sending process in an ancillary message\&. Defaults to
708 \fBfalse\fR\&.
709 .RE
710 .PP
711 \fIPassPIDFD=\fR
712 .RS 4
713 Takes a boolean value\&. This controls the
714 \fBSO_PASSPIDFD\fR
715 socket option, which allows
716 \fBAF_UNIX\fR
717 sockets to receive the pidfd of the sending process in an ancillary message\&. Defaults to
718 \fBfalse\fR\&.
719 .sp
720 Added in version 258\&.
721 .RE
722 .PP
723 \fIPassSecurity=\fR
724 .RS 4
725 Takes a boolean value\&. This controls the
726 \fBSO_PASSSEC\fR
727 socket option, which allows
728 \fBAF_UNIX\fR
729 sockets to receive the security context of the sending process in an ancillary message\&. Defaults to
730 \fBfalse\fR\&.
731 .RE
732 .PP
733 \fIPassPacketInfo=\fR
734 .RS 4
735 Takes a boolean value\&. This controls the
736 \fBIP_PKTINFO\fR,
737 \fBIPV6_RECVPKTINFO\fR,
738 \fBNETLINK_PKTINFO\fR
739 or
740 \fBPACKET_AUXDATA\fR
741 socket options, which enable reception of additional per\-packet metadata as ancillary message, on
742 \fBAF_INET\fR,
743 \fBAF_INET6\fR,
744 \fBAF_UNIX\fR
745 and
746 \fBAF_PACKET\fR
747 sockets\&. Defaults to
748 \fBfalse\fR\&.
749 .sp
750 Added in version 246\&.
751 .RE
752 .PP
753 \fIAcceptFileDescriptors=\fR
754 .RS 4
755 Takes a boolean value\&. This controls the
756 \fBSO_PASSRIGHTS\fR
757 socket option, which when disabled prohibits the peer from sending
758 \fBSCM_RIGHTS\fR
759 ancillary messages (aka file descriptors) via
760 \fBAF_UNIX\fR
761 sockets\&. Defaults to
762 \fBtrue\fR\&.
763 .sp
764 Added in version 258\&.
765 .RE
766 .PP
767 \fITimestamping=\fR
768 .RS 4
769 Takes one of
770 "off",
771 "us"
772 (alias:
773 "usec",
774 "μs") or
775 "ns"
776 (alias:
777 "nsec")\&. This controls the
778 \fBSO_TIMESTAMP\fR
779 or
780 \fBSO_TIMESTAMPNS\fR
781 socket options, and enables whether ingress network traffic shall carry timestamping metadata\&. Defaults to
782 \fBoff\fR\&.
783 .sp
784 Added in version 247\&.
785 .RE
786 .PP
787 \fITCPCongestion=\fR
788 .RS 4
789 Takes a string value\&. Controls the TCP congestion algorithm used by this socket\&. Should be one of
790 "westwood",
791 "reno",
792 "cubic",
793 "lp"
794 or any other available algorithm supported by the IP stack\&. This setting applies only to stream sockets\&.
795 .RE
796 .PP
797 \fIExecStartPre=\fR, \fIExecStartPost=\fR
798 .RS 4
799 Takes one or more command lines, which are executed before or after the listening sockets/FIFOs are created and bound, respectively\&. The first token of the command line must be an absolute filename, then followed by arguments for the process\&. Multiple command lines may be specified following the same scheme as used for
800 \fIExecStartPre=\fR
801 of service unit files\&.
802 .RE
803 .PP
804 \fIExecStopPre=\fR, \fIExecStopPost=\fR
805 .RS 4
806 Additional commands that are executed before or after the listening sockets/FIFOs are closed and removed, respectively\&. Multiple command lines may be specified following the same scheme as used for
807 \fIExecStartPre=\fR
808 of service unit files\&.
809 .RE
810 .PP
811 \fITimeoutSec=\fR
812 .RS 4
813 Configures the time to wait for the commands specified in
814 \fIExecStartPre=\fR,
815 \fIExecStartPost=\fR,
816 \fIExecStopPre=\fR
817 and
818 \fIExecStopPost=\fR
819 to finish\&. If a command does not exit within the configured time, the socket will be considered failed and be shut down again\&. All commands still running will be terminated forcibly via
820 \fBSIGTERM\fR, and after another delay of this time with
821 \fBSIGKILL\fR\&. (See
822 \fBKillMode=\fR
823 in
824 \fBsystemd.kill\fR(5)\&.) Takes a unit\-less value in seconds, or a time span value such as "5min 20s"\&. Pass
825 "0"
826 to disable the timeout logic\&. Defaults to
827 \fIDefaultTimeoutStartSec=\fR
828 from the manager configuration file (see
829 \fBsystemd-system.conf\fR(5))\&.
830 .RE
831 .PP
832 \fIService=\fR
833 .RS 4
834 Specifies the service unit name to activate on incoming traffic\&. This setting is only allowed for sockets with
835 \fIAccept=no\fR\&. It defaults to the service that bears the same name as the socket (with the suffix replaced)\&. In most cases, it should not be necessary to use this option\&. Note that setting this parameter might result in additional dependencies to be added to the unit (see above)\&.
836 .RE
837 .PP
838 \fIRemoveOnStop=\fR
839 .RS 4
840 Takes a boolean argument\&. If enabled, any file nodes created by this socket unit are removed when it is stopped\&. This applies to
841 \fBAF_UNIX\fR
842 sockets in the file system, POSIX message queues, FIFOs, as well as any symlinks to them configured with
843 \fISymlinks=\fR\&. Normally, it should not be necessary to use this option, and is not recommended as services might continue to run after the socket unit has been terminated and it should still be possible to communicate with them via their file system node\&. Defaults to off\&.
844 .sp
845 Added in version 214\&.
846 .RE
847 .PP
848 \fISymlinks=\fR
849 .RS 4
850 Takes a list of file system paths\&. The specified paths will be created as symlinks to the
851 \fBAF_UNIX\fR
852 socket path or FIFO path of this socket unit\&. If this setting is used, only one
853 \fBAF_UNIX\fR
854 socket in the file system or one FIFO may be configured for the socket unit\&. Use this option to manage one or more symlinked alias names for a socket, binding their lifecycle together\&. Note that if creation of a symlink fails this is not considered fatal for the socket unit, and the socket unit may still start\&. If an empty string is assigned, the list of paths is reset\&. Defaults to an empty list\&.
855 .sp
856 Added in version 214\&.
857 .RE
858 .PP
859 \fIFileDescriptorName=\fR
860 .RS 4
861 Assigns a name to all file descriptors this socket unit encapsulates\&. This is useful to help activated services identify specific file descriptors, if multiple fds are passed\&. Services may use the
862 \fBsd_listen_fds_with_names\fR(3)
863 call to acquire the names configured for the received file descriptors\&. Names may contain any ASCII character, but must exclude control characters and
864 ":", and must be at most 255 characters in length\&. If this setting is not used, the file descriptor name defaults to the name of the socket unit (including its
865 \&.socket
866 suffix) when
867 \fIAccept=no\fR,
868 "connection"
869 otherwise\&.
870 .sp
871 Added in version 227\&.
872 .RE
873 .PP
874 \fITriggerLimitIntervalSec=\fR, \fITriggerLimitBurst=\fR
875 .RS 4
876 Configures a limit on how often this socket unit may be activated within a specific time interval\&. The
877 \fITriggerLimitIntervalSec=\fR
878 setting may be used to configure the length of the time interval in the usual time units
879 "us",
880 "ms",
881 "s",
882 "min",
883 "h", \&... and defaults to 2s (See
884 \fBsystemd.time\fR(7)
885 for details on the various time units understood)\&. The
886 \fITriggerLimitBurst=\fR
887 setting takes a positive integer value and specifies the number of permitted activations per time interval, and defaults to 200 for
888 \fIAccept=yes\fR
889 sockets (thus by default permitting 200 activations per 2s), and 20 otherwise (20 activations per 2s)\&. Set either to 0 to disable any form of trigger rate limiting\&.
890 .sp
891 If the limit is hit, the socket unit is placed into a failure mode, and will not be connectible anymore until restarted\&. Note that this limit is enforced before the service activation is enqueued\&.
892 .sp
893 Compare with
894 \fIPollLimitIntervalSec=\fR/\fIPollLimitBurst=\fR
895 described below, which implements a temporary slowdown if a socket unit is flooded with incoming traffic, as opposed to the permanent failure state
896 \fITriggerLimitIntervalSec=\fR/\fITriggerLimitBurst=\fR
897 results in\&.
898 .sp
899 Added in version 230\&.
900 .RE
901 .PP
902 \fIPollLimitIntervalSec=\fR, \fIPollLimitBurst=\fR
903 .RS 4
904 Configures a limit on how often polling events on the file descriptors backing this socket unit will be considered\&. This pair of settings is similar to
905 \fITriggerLimitIntervalSec=\fR/\fITriggerLimitBurst=\fR
906 but instead of putting a (fatal) limit on the activation frequency puts a (transient) limit on the polling frequency\&. The expected parameter syntax and range are identical to that of the aforementioned options, and can be disabled the same way\&.
907 .sp
908 If the polling limit is hit polling is temporarily disabled on it until the specified time window passes\&. The polling limit hence slows down connection attempts if hit, but unlike the trigger limit will not cause permanent failures\&. It\*(Aqs the recommended mechanism to deal with DoS attempts through packet flooding\&.
909 .sp
910 The polling limit is enforced per file descriptor to listen on, as opposed to the trigger limit which is enforced for the entire socket unit\&. This distinction matters for socket units that listen on multiple file descriptors (i\&.e\&. have multiple
911 \fIListenXYZ=\fR
912 stanzas)\&.
913 .sp
914 These setting defaults to 150 (in case of
915 \fIAccept=yes\fR) and 15 (otherwise) polling events per 2s\&. This is considerably lower than the default values for the trigger limit (see above) and means that the polling limit should typically ensure the trigger limit is never hit, unless one of them is reconfigured or disabled\&.
916 .sp
917 Added in version 255\&.
918 .RE
919 .PP
920 \fIDeferTrigger=\fR
921 .RS 4
922 Takes a boolean argument, or
923 "patient"\&. May only be used when
924 \fIAccept=no\fR\&. If enabled, job mode
925 "lenient"
926 instead of
927 "replace"
928 is used when triggering the service, which means currently activating/running units that conflict with the service won\*(Aqt be disturbed/brought down\&. Furthermore, if a conflict exists, the socket unit will wait for current job queue to complete and potentially defer the activation by then\&. An upper limit of total time to wait can be configured via
929 \fIDeferTriggerMaxSec=\fR\&. If set to
930 \fByes\fR, the socket unit will fail if all jobs have finished or the timeout has been reached but the conflict remains\&. If
931 \fBpatient\fR, always wait until
932 \fIDeferTriggerMaxSec=\fR
933 elapses\&. Defaults to no\&.
934 .sp
935 This setting is particularly useful if the socket unit should stay active across switch\-root/soft\-reboot operations while the triggered service is stopped\&.
936 .sp
937 Added in version 258\&.
938 .RE
939 .PP
940 \fIDeferTriggerMaxSec=\fR
941 .RS 4
942 Configures the maximum time to defer the triggering when
943 \fIDeferTrigger=\fR
944 is enabled\&. If the service cannot be activated within the specified time, the socket will be considered failed and get terminated\&. Takes a unit\-less value in seconds, or a time span value such as "5min 20s"\&. Pass
945 "0"
946 or
947 "infinity"
948 to disable the timeout logic (the default)\&.
949 .sp
950 Added in version 258\&.
951 .RE
952 .PP
953 \fIPassFileDescriptorsToExec=\fR
954 .RS 4
955 Takes a boolean argument\&. Defaults to off\&. If enabled, file descriptors created by the socket unit are passed to
956 \fIExecStartPost=\fR,
957 \fIExecStopPre=\fR, and
958 \fIExecStopPost=\fR
959 commands from the socket unit\&. The passed file descriptors can be accessed with
960 \fBsd_listen_fds\fR(3)
961 as if the commands were invoked from the associated service units\&. Note that
962 \fIExecStartPre=\fR
963 command cannot access socket file descriptors\&.
964 .sp
965 Added in version 256\&.
966 .RE
967 .PP
968 Check
969 \fBsystemd.unit\fR(5),
970 \fBsystemd.exec\fR(5), and
971 \fBsystemd.kill\fR(5)
972 for more settings\&.
973 .SH "SEE ALSO"
974 .PP
975 \fBsystemd\fR(1), \fBsystemctl\fR(1), \fBsystemd-system.conf\fR(5), \fBsystemd.unit\fR(5), \fBsystemd.exec\fR(5), \fBsystemd.kill\fR(5), \fBsystemd.resource-control\fR(5), \fBsystemd.service\fR(5), \fBsystemd.directives\fR(7), \fBsd_listen_fds\fR(3), \fBsd_listen_fds_with_names\fR(3)
976 .PP
977 For more extensive descriptions see the "systemd for Developers" series:
978 \m[blue]\fBSocket Activation\fR\m[]\&\s-2\u[5]\d\s+2, \m[blue]\fBSocket Activation, part II\fR\m[]\&\s-2\u[6]\d\s+2, \m[blue]\fBConverting inetd Services\fR\m[]\&\s-2\u[7]\d\s+2, \m[blue]\fBSocket Activated Internet Services and OS Containers\fR\m[]\&\s-2\u[8]\d\s+2\&.
979 .SH "NOTES"
980 .IP " 1." 4
981 USB FunctionFS
982 .RS 4
983 \%https://docs.kernel.org/usb/functionfs.html
984 .RE
985 .IP " 2." 4
986 RFC 3875
987 .RS 4
988 \%https://datatracker.ietf.org/doc/html/rfc3875
989 .RE
990 .IP " 3." 4
991 TCP Keepalive HOWTO
992 .RS 4
993 \%http://www.tldp.org/HOWTO/html_single/TCP-Keepalive-HOWTO/
994 .RE
995 .IP " 4." 4
996 Smack
997 .RS 4
998 \%https://docs.kernel.org/admin-guide/LSM/Smack.html
999 .RE
1000 .IP " 5." 4
1001 Socket Activation
1002 .RS 4
1003 \%https://0pointer.de/blog/projects/socket-activation.html
1004 .RE
1005 .IP " 6." 4
1006 Socket Activation, part II
1007 .RS 4
1008 \%https://0pointer.de/blog/projects/socket-activation2.html
1009 .RE
1010 .IP " 7." 4
1011 Converting inetd Services
1012 .RS 4
1013 \%https://0pointer.de/blog/projects/inetd.html
1014 .RE
1015 .IP " 8." 4
1016 Socket Activated Internet Services and OS Containers
1017 .RS 4
1018 \%https://0pointer.de/blog/projects/socket-activated-containers.html
1019 .RE