1SYSTEMD.SOCKET(5) systemd.socket SYSTEMD.SOCKET(5)
2
6 systemd.socket - Socket unit configuration
7
9 socket.socket
10
12 A unit configuration file whose name ends in ".socket" encodes
13 information about an IPC or network socket or a file system FIFO
14 controlled and supervised by systemd, for socket-based activation.
15 This man page lists the configuration options specific to this unit
17 type. See systemd.unit(5) for the common options of all unit
18 configuration files. The common configuration items are configured in
19 the generic [Unit] and [Install] sections. The socket specific
20 configuration options are configured in the [Socket] section.
21 Additional options are listed in systemd.exec(5), which define the
23 execution environment the ExecStartPre=, ExecStartPost=, ExecStopPre=
24 and ExecStopPost= commands are executed in, and in systemd.kill(5),
25 which define the way the processes are terminated, and in
26 systemd.resource-control(5), which configure resource control settings
27 for the processes of the socket.
28 For each socket unit, a matching service unit must exist, describing
30 the service to start on incoming traffic on the socket (see
31 systemd.service(5) for more information about .service units). The name
32 of the .service unit is by default the same as the name of the .socket
33 unit, but can be altered with the Service= option described below.
34 Depending on the setting of the Accept= option described below, this
35 .service unit must either be named like the .socket unit, but with the
36 suffix replaced, unless overridden with Service=; or it must be a
37 template unit named the same way. Example: a socket file foo.socket
38 needs a matching service foo.service if Accept=no is set. If Accept=yes
39 is set, a service template foo@.service must exist from which services
40 are instantiated for each incoming connection.
41 No implicit WantedBy= or RequiredBy= dependency from the socket to the
43 service is added. This means that the service may be started without
44 the socket, in which case it must be able to open sockets by itself. To
45 prevent this, an explicit Requires= dependency may be added.
46 Socket units may be used to implement on-demand starting of services,
48 as well as parallelized starting of services. See the blog stories
49 linked at the end for an introduction.
50 Note that the daemon software configured for socket activation with
52 socket units needs to be able to accept sockets from systemd, either
53 via systemd's native socket passing interface (see sd_listen_fds(3) for
54 details about the precise protocol used and the order in which the file
55 descriptors are passed) or via traditional inetd(8)-style socket
56 passing (i.e. sockets passed in via standard input and output, using
57 StandardInput=socket in the service file).
58 All network sockets allocated through .socket units are allocated in
60 the host's network namespace (see network_namespaces(7)). This does not
61 mean however that the service activated by a configured socket unit has
62 to be part of the host's network namespace as well. It is supported and
63 even good practice to run services in their own network namespace (for
64 example through PrivateNetwork=, see systemd.exec(5)), receiving only
65 the sockets configured through socket-activation from the host's
66 namespace. In such a set-up communication within the host's network
67 namespace is only permitted through the activation sockets passed in
68 while all sockets allocated from the service code itself will be
69 associated with the service's own namespace, and thus possibly subject
70 to a restrictive configuration.
71
73 Implicit Dependencies
74 The following dependencies are implicitly added:
75 • Socket units automatically gain a Before= dependency on the service
77 units they activate.
78 • Socket units referring to file system paths (such as AF_UNIX
80 sockets or FIFOs) implicitly gain Requires= and After= dependencies
81 on all mount units necessary to access those paths.
82 • Socket units using the BindToDevice= setting automatically gain a
84 BindsTo= and After= dependency on the device unit encapsulating the
85 specified network interface.
86 Additional implicit dependencies may be added as result of execution
88 and resource control parameters as documented in systemd.exec(5) and
89 systemd.resource-control(5).
90 Default Dependencies
92 The following dependencies are added unless DefaultDependencies=no is
93 set:
94 • Socket units automatically gain a Before= dependency on
96 sockets.target.
97 • Socket units automatically gain a pair of After= and Requires=
99 dependency on sysinit.target, and a pair of Before= and Conflicts=
100 dependencies on shutdown.target. These dependencies ensure that the
101 socket unit is started before normal services at boot, and is
102 stopped on shutdown. Only sockets involved with early boot or late
103 system shutdown should disable DefaultDependencies= option.
104
106 Socket unit files may include [Unit] and [Install] sections, which are
107 described in systemd.unit(5).
108 Socket unit files must include a [Socket] section, which carries
110 information about the socket or FIFO it supervises. A number of options
111 that may be used in this section are shared with other unit types.
112 These options are documented in systemd.exec(5) and systemd.kill(5).
113 The options specific to the [Socket] section of socket units are the
114 following:
115 ListenStream=, ListenDatagram=, ListenSequentialPacket=
117 Specifies an address to listen on for a stream (SOCK_STREAM),
118 datagram (SOCK_DGRAM), or sequential packet (SOCK_SEQPACKET)
119 socket, respectively. The address can be written in various
120 formats:
121 If the address starts with a slash ("/"), it is read as file system
123 socket in the AF_UNIX socket family.
124 If the address starts with an at symbol ("@"), it is read as
126 abstract namespace socket in the AF_UNIX family. The "@" is
127 replaced with a NUL character before binding. For details, see
128 unix(7).
129 If the address string is a single number, it is read as port number
131 to listen on via IPv6. Depending on the value of BindIPv6Only= (see
132 below) this might result in the service being available via both
133 IPv6 and IPv4 (default) or just via IPv6.
134 If the address string is a string in the format "v.w.x.y:z", it is
136 interpreted as IPv4 address v.w.x.y and port z.
137 If the address string is a string in the format "[x]:y", it is
139 interpreted as IPv6 address x and port y. An optional interface
140 scope (interface name or number) may be specified after a "%"
141 symbol: "[x]:y%dev". Interface scopes are only useful with
142 link-local addresses, because the kernel ignores them in other
143 cases. Note that if an address is specified as IPv6, it might still
144 make the service available via IPv4 too, depending on the
145 BindIPv6Only= setting (see below).
146 If the address string is a string in the format "vsock:x:y", it is
148 read as CID x on a port y address in the AF_VSOCK family. The CID
149 is a unique 32-bit integer identifier in AF_VSOCK analogous to an
150 IP address. Specifying the CID is optional, and may be set to the
151 empty string.
152 Note that SOCK_SEQPACKET (i.e. ListenSequentialPacket=) is only
154 available for AF_UNIX sockets. SOCK_STREAM (i.e. ListenStream=)
155 when used for IP sockets refers to TCP sockets, SOCK_DGRAM (i.e.
156 ListenDatagram=) to UDP.
157 These options may be specified more than once, in which case
159 incoming traffic on any of the sockets will trigger service
160 activation, and all listed sockets will be passed to the service,
161 regardless of whether there is incoming traffic on them or not. If
162 the empty string is assigned to any of these options, the list of
163 addresses to listen on is reset, all prior uses of any of these
164 options will have no effect.
165 It is also possible to have more than one socket unit for the same
167 service when using Service=, and the service will receive all the
168 sockets configured in all the socket units. Sockets configured in
169 one unit are passed in the order of configuration, but no ordering
170 between socket units is specified.
171 If an IP address is used here, it is often desirable to listen on
173 it before the interface it is configured on is up and running, and
174 even regardless of whether it will be up and running at any point.
175 To deal with this, it is recommended to set the FreeBind= option
176 described below.
177 ListenFIFO=
179 Specifies a file system FIFO (see fifo(7) for details) to listen
180 on. This expects an absolute file system path as argument. Behavior
181 otherwise is very similar to the ListenDatagram= directive above.
182 ListenSpecial=
184 Specifies a special file in the file system to listen on. This
185 expects an absolute file system path as argument. Behavior
186 otherwise is very similar to the ListenFIFO= directive above. Use
187 this to open character device nodes as well as special files in
188 /proc/ and /sys/.
189 ListenNetlink=
191 Specifies a Netlink family to create a socket for to listen on.
192 This expects a short string referring to the AF_NETLINK family name
193 (such as audit or kobject-uevent) as argument, optionally suffixed
194 by a whitespace followed by a multicast group integer. Behavior
195 otherwise is very similar to the ListenDatagram= directive above.
196 ListenMessageQueue=
198 Specifies a POSIX message queue name to listen on (see
199 mq_overview(7) for details). This expects a valid message queue
200 name (i.e. beginning with "/"). Behavior otherwise is very similar
201 to the ListenFIFO= directive above. On Linux message queue
202 descriptors are actually file descriptors and can be inherited
203 between processes.
204 ListenUSBFunction=
206 Specifies a USB FunctionFS[1] endpoints location to listen on, for
207 implementation of USB gadget functions. This expects an absolute
208 file system path of a FunctionFS mount point as the argument.
209 Behavior otherwise is very similar to the ListenFIFO= directive
210 above. Use this to open the FunctionFS endpoint ep0. When using
211 this option, the activated service has to have the
212 USBFunctionDescriptors= and USBFunctionStrings= options set.
213 SocketProtocol=
215 Takes one of udplite or sctp. The socket will use the UDP-Lite
216 (IPPROTO_UDPLITE) or SCTP (IPPROTO_SCTP) protocol, respectively.
217 BindIPv6Only=
219 Takes one of default, both or ipv6-only. Controls the IPV6_V6ONLY
220 socket option (see ipv6(7) for details). If both, IPv6 sockets
221 bound will be accessible via both IPv4 and IPv6. If ipv6-only, they
222 will be accessible via IPv6 only. If default (which is the default,
223 surprise!), the system wide default setting is used, as controlled
224 by /proc/sys/net/ipv6/bindv6only, which in turn defaults to the
225 equivalent of both.
226 Backlog=
228 Takes an unsigned 32-bit integer argument. Specifies the number of
229 connections to queue that have not been accepted yet. This setting
230 matters only for stream and sequential packet sockets. See
231 listen(2) for details. Note that this value is silently capped by
232 the "net.core.somaxconn" sysctl, which typically defaults to 4096.
233 By default this is set to 4294967295, so that the sysctl takes full
234 effect.
235 BindToDevice=
237 Specifies a network interface name to bind this socket to. If set,
238 traffic will only be accepted from the specified network
239 interfaces. This controls the SO_BINDTODEVICE socket option (see
240 socket(7) for details). If this option is used, an implicit
241 dependency from this socket unit on the network interface device
242 unit is created (see systemd.device(5)). Note that setting this
243 parameter might result in additional dependencies to be added to
244 the unit (see above).
245 SocketUser=, SocketGroup=
247 Takes a UNIX user/group name. When specified, all AF_UNIX sockets
248 and FIFO nodes in the file system are owned by the specified user
249 and group. If unset (the default), the nodes are owned by the root
250 user/group (if run in system context) or the invoking user/group
251 (if run in user context). If only a user is specified but no group,
252 then the group is derived from the user's default group.
253 SocketMode=
255 If listening on a file system socket or FIFO, this option specifies
256 the file system access mode used when creating the file node. Takes
257 an access mode in octal notation. Defaults to 0666.
258 DirectoryMode=
260 If listening on a file system socket or FIFO, the parent
261 directories are automatically created if needed. This option
262 specifies the file system access mode used when creating these
263 directories. Takes an access mode in octal notation. Defaults to
264 0755.
265 Accept=
267 Takes a boolean argument. If yes, a service instance is spawned for
268 each incoming connection and only the connection socket is passed
269 to it. If no, all listening sockets themselves are passed to the
270 started service unit, and only one service unit is spawned for all
271 connections (also see above). This value is ignored for datagram
272 sockets and FIFOs where a single service unit unconditionally
273 handles all incoming traffic. Defaults to no. For performance
274 reasons, it is recommended to write new daemons only in a way that
275 is suitable for Accept=no. A daemon listening on an AF_UNIX socket
276 may, but does not need to, call close(2) on the received socket
277 before exiting. However, it must not unlink the socket from a file
278 system. It should not invoke shutdown(2) on sockets it got with
279 Accept=no, but it may do so for sockets it got with Accept=yes set.
280 Setting Accept=yes is mostly useful to allow daemons designed for
281 usage with inetd(8) to work unmodified with systemd socket
282 activation.
283 Note that depending on this setting the services activated by units
285 of this type are either regular services (in case of Accept=no) or
286 instances of templated services (in case of Accept=yes). See the
287 Description section above for a more detailed discussion of the
288 naming rules of triggered services.
289 For IPv4 and IPv6 connections, the REMOTE_ADDR environment variable
291 will contain the remote IP address, and REMOTE_PORT will contain
292 the remote port. This is the same as the format used by CGI. For
293 SOCK_RAW, the port is the IP protocol.
294 It is recommended to set CollectMode=inactive-or-failed for service
296 instances activated via Accept=yes, to ensure that failed
297 connection services are cleaned up and released from memory, and do
298 not accumulate.
299 Writable=
301 Takes a boolean argument. May only be used in conjunction with
302 ListenSpecial=. If true, the specified special file is opened in
303 read-write mode, if false, in read-only mode. Defaults to false.
304 FlushPending=
306 Takes a boolean argument. May only be used when Accept=no. If yes,
307 the socket's buffers are cleared after the triggered service
308 exited. This causes any pending data to be flushed and any pending
309 incoming connections to be rejected. If no, the socket's buffers
310 won't be cleared, permitting the service to handle any pending
311 connections after restart, which is the usually expected behaviour.
312 Defaults to no.
313 MaxConnections=
315 The maximum number of connections to simultaneously run services
316 instances for, when Accept=yes is set. If more concurrent
317 connections are coming in, they will be refused until at least one
318 existing connection is terminated. This setting has no effect on
319 sockets configured with Accept=no or datagram sockets. Defaults to
320 64.
321 MaxConnectionsPerSource=
323 The maximum number of connections for a service per source IP
324 address. This is very similar to the MaxConnections= directive
325 above. Disabled by default.
326 KeepAlive=
328 Takes a boolean argument. If true, the TCP/IP stack will send a
329 keep alive message after 2h (depending on the configuration of
330 /proc/sys/net/ipv4/tcp_keepalive_time) for all TCP streams accepted
331 on this socket. This controls the SO_KEEPALIVE socket option (see
332 socket(7) and the TCP Keepalive HOWTO[2] for details.) Defaults to
333 false.
334 KeepAliveTimeSec=
336 Takes time (in seconds) as argument. The connection needs to remain
337 idle before TCP starts sending keepalive probes. This controls the
338 TCP_KEEPIDLE socket option (see socket(7) and the TCP Keepalive
339 HOWTO[2] for details.) Default value is 7200 seconds (2 hours).
340 KeepAliveIntervalSec=
342 Takes time (in seconds) as argument between individual keepalive
343 probes, if the socket option SO_KEEPALIVE has been set on this
344 socket. This controls the TCP_KEEPINTVL socket option (see
345 socket(7) and the TCP Keepalive HOWTO[2] for details.) Default
346 value is 75 seconds.
347 KeepAliveProbes=
349 Takes an integer as argument. It is the number of unacknowledged
350 probes to send before considering the connection dead and notifying
351 the application layer. This controls the TCP_KEEPCNT socket option
352 (see socket(7) and the TCP Keepalive HOWTO[2] for details.) Default
353 value is 9.
354 NoDelay=
356 Takes a boolean argument. TCP Nagle's algorithm works by combining
357 a number of small outgoing messages, and sending them all at once.
358 This controls the TCP_NODELAY socket option (see tcp(7)). Defaults
359 to false.
360 Priority=
362 Takes an integer argument controlling the priority for all traffic
363 sent from this socket. This controls the SO_PRIORITY socket option
364 (see socket(7) for details.).
365 DeferAcceptSec=
367 Takes time (in seconds) as argument. If set, the listening process
368 will be awakened only when data arrives on the socket, and not
369 immediately when connection is established. When this option is
370 set, the TCP_DEFER_ACCEPT socket option will be used (see tcp(7)),
371 and the kernel will ignore initial ACK packets without any data.
372 The argument specifies the approximate amount of time the kernel
373 should wait for incoming data before falling back to the normal
374 behavior of honoring empty ACK packets. This option is beneficial
375 for protocols where the client sends the data first (e.g. HTTP, in
376 contrast to SMTP), because the server process will not be woken up
377 unnecessarily before it can take any action.
378 If the client also uses the TCP_DEFER_ACCEPT option, the latency of
380 the initial connection may be reduced, because the kernel will send
381 data in the final packet establishing the connection (the third
382 packet in the "three-way handshake").
383 Disabled by default.
385 ReceiveBuffer=, SendBuffer=
387 Takes an integer argument controlling the receive or send buffer
388 sizes of this socket, respectively. This controls the SO_RCVBUF and
389 SO_SNDBUF socket options (see socket(7) for details.). The usual
390 suffixes K, M, G are supported and are understood to the base of
391 1024.
392 IPTOS=
394 Takes an integer argument controlling the IP Type-Of-Service field
395 for packets generated from this socket. This controls the IP_TOS
396 socket option (see ip(7) for details.). Either a numeric string or
397 one of low-delay, throughput, reliability or low-cost may be
398 specified.
399 IPTTL=
401 Takes an integer argument controlling the IPv4 Time-To-Live/IPv6
402 Hop-Count field for packets generated from this socket. This sets
403 the IP_TTL/IPV6_UNICAST_HOPS socket options (see ip(7) and ipv6(7)
404 for details.)
405 Mark=
407 Takes an integer value. Controls the firewall mark of packets
408 generated by this socket. This can be used in the firewall logic to
409 filter packets from this socket. This sets the SO_MARK socket
410 option. See iptables(8) for details.
411 ReusePort=
413 Takes a boolean value. If true, allows multiple bind(2)s to this
414 TCP or UDP port. This controls the SO_REUSEPORT socket option. See
415 socket(7) for details.
416 SmackLabel=, SmackLabelIPIn=, SmackLabelIPOut=
418 Takes a string value. Controls the extended attributes
419 "security.SMACK64", "security.SMACK64IPIN" and
420 "security.SMACK64IPOUT", respectively, i.e. the security label of
421 the FIFO, or the security label for the incoming or outgoing
422 connections of the socket, respectively. See Smack[3] for details.
423 SELinuxContextFromNet=
425 Takes a boolean argument. When true, systemd will attempt to figure
426 out the SELinux label used for the instantiated service from the
427 information handed by the peer over the network. Note that only the
428 security level is used from the information provided by the peer.
429 Other parts of the resulting SELinux context originate from either
430 the target binary that is effectively triggered by socket unit or
431 from the value of the SELinuxContext= option. This configuration
432 option applies only when activated service is passed in single
433 socket file descriptor, i.e. service instances that have standard
434 input connected to a socket or services triggered by exactly one
435 socket unit. Also note that this option is useful only when MLS/MCS
436 SELinux policy is deployed. Defaults to "false".
437 PipeSize=
439 Takes a size in bytes. Controls the pipe buffer size of FIFOs
440 configured in this socket unit. See fcntl(2) for details. The usual
441 suffixes K, M, G are supported and are understood to the base of
442 1024.
443 MessageQueueMaxMessages=, MessageQueueMessageSize=
445 These two settings take integer values and control the mq_maxmsg
446 field or the mq_msgsize field, respectively, when creating the
447 message queue. Note that either none or both of these variables
448 need to be set. See mq_setattr(3) for details.
449 FreeBind=
451 Takes a boolean value. Controls whether the socket can be bound to
452 non-local IP addresses. This is useful to configure sockets
453 listening on specific IP addresses before those IP addresses are
454 successfully configured on a network interface. This sets the
455 IP_FREEBIND/IPV6_FREEBIND socket option. For robustness reasons it
456 is recommended to use this option whenever you bind a socket to a
457 specific IP address. Defaults to false.
458 Transparent=
460 Takes a boolean value. Controls the IP_TRANSPARENT/IPV6_TRANSPARENT
461 socket option. Defaults to false.
462 Broadcast=
464 Takes a boolean value. This controls the SO_BROADCAST socket
465 option, which allows broadcast datagrams to be sent from this
466 socket. Defaults to false.
467 PassCredentials=
469 Takes a boolean value. This controls the SO_PASSCRED socket option,
470 which allows AF_UNIX sockets to receive the credentials of the
471 sending process in an ancillary message. Defaults to false.
472 PassSecurity=
474 Takes a boolean value. This controls the SO_PASSSEC socket option,
475 which allows AF_UNIX sockets to receive the security context of the
476 sending process in an ancillary message. Defaults to false.
477 PassPacketInfo=
479 Takes a boolean value. This controls the IP_PKTINFO,
480 IPV6_RECVPKTINFO, NETLINK_PKTINFO or PACKET_AUXDATA socket options,
481 which enable reception of additional per-packet metadata as
482 ancillary message, on AF_INET, AF_INET6, AF_UNIX and AF_PACKET
483 sockets. Defaults to false.
484 Timestamping=
486 Takes one of "off", "us" (alias: "usec", "μs") or "ns" (alias:
487 "nsec"). This controls the SO_TIMESTAMP or SO_TIMESTAMPNS socket
488 options, and enables whether ingress network traffic shall carry
489 timestamping metadata. Defaults to off.
490 TCPCongestion=
492 Takes a string value. Controls the TCP congestion algorithm used by
493 this socket. Should be one of "westwood", "veno", "cubic", "lp" or
494 any other available algorithm supported by the IP stack. This
495 setting applies only to stream sockets.
496 ExecStartPre=, ExecStartPost=
498 Takes one or more command lines, which are executed before or after
499 the listening sockets/FIFOs are created and bound, respectively.
500 The first token of the command line must be an absolute filename,
501 then followed by arguments for the process. Multiple command lines
502 may be specified following the same scheme as used for
503 ExecStartPre= of service unit files.
504 ExecStopPre=, ExecStopPost=
506 Additional commands that are executed before or after the listening
507 sockets/FIFOs are closed and removed, respectively. Multiple
508 command lines may be specified following the same scheme as used
509 for ExecStartPre= of service unit files.
510 TimeoutSec=
512 Configures the time to wait for the commands specified in
513 ExecStartPre=, ExecStartPost=, ExecStopPre= and ExecStopPost= to
514 finish. If a command does not exit within the configured time, the
515 socket will be considered failed and be shut down again. All
516 commands still running will be terminated forcibly via SIGTERM, and
517 after another delay of this time with SIGKILL. (See KillMode= in
518 systemd.kill(5).) Takes a unit-less value in seconds, or a time
519 span value such as "5min 20s". Pass "0" to disable the timeout
520 logic. Defaults to DefaultTimeoutStartSec= from the manager
521 configuration file (see systemd-system.conf(5)).
522 Service=
524 Specifies the service unit name to activate on incoming traffic.
525 This setting is only allowed for sockets with Accept=no. It
526 defaults to the service that bears the same name as the socket
527 (with the suffix replaced). In most cases, it should not be
528 necessary to use this option. Note that setting this parameter
529 might result in additional dependencies to be added to the unit
530 (see above).
531 RemoveOnStop=
533 Takes a boolean argument. If enabled, any file nodes created by
534 this socket unit are removed when it is stopped. This applies to
535 AF_UNIX sockets in the file system, POSIX message queues, FIFOs, as
536 well as any symlinks to them configured with Symlinks=. Normally,
537 it should not be necessary to use this option, and is not
538 recommended as services might continue to run after the socket unit
539 has been terminated and it should still be possible to communicate
540 with them via their file system node. Defaults to off.
541 Symlinks=
543 Takes a list of file system paths. The specified paths will be
544 created as symlinks to the AF_UNIX socket path or FIFO path of this
545 socket unit. If this setting is used, only one AF_UNIX socket in
546 the file system or one FIFO may be configured for the socket unit.
547 Use this option to manage one or more symlinked alias names for a
548 socket, binding their lifecycle together. Note that if creation of
549 a symlink fails this is not considered fatal for the socket unit,
550 and the socket unit may still start. If an empty string is
551 assigned, the list of paths is reset. Defaults to an empty list.
552 FileDescriptorName=
554 Assigns a name to all file descriptors this socket unit
555 encapsulates. This is useful to help activated services identify
556 specific file descriptors, if multiple fds are passed. Services may
557 use the sd_listen_fds_with_names(3) call to acquire the names
558 configured for the received file descriptors. Names may contain any
559 ASCII character, but must exclude control characters and ":", and
560 must be at most 255 characters in length. If this setting is not
561 used, the file descriptor name defaults to the name of the socket
562 unit, including its .socket suffix.
563 TriggerLimitIntervalSec=, TriggerLimitBurst=
565 Configures a limit on how often this socket unit may be activated
566 within a specific time interval. The TriggerLimitIntervalSec=
567 setting may be used to configure the length of the time interval in
568 the usual time units "us", "ms", "s", "min", "h", ... and defaults
569 to 2s (See systemd.time(7) for details on the various time units
570 understood). The TriggerLimitBurst= setting takes a positive
571 integer value and specifies the number of permitted activations per
572 time interval, and defaults to 200 for Accept=yes sockets (thus by
573 default permitting 200 activations per 2s), and 20 otherwise (20
574 activations per 2s). Set either to 0 to disable any form of trigger
575 rate limiting.
576 If the limit is hit, the socket unit is placed into a failure mode,
578 and will not be connectible anymore until restarted. Note that this
579 limit is enforced before the service activation is enqueued.
580 Compare with PollLimitIntervalSec=/PollLimitBurst= described below,
582 which implements a temporary slowdown if a socket unit is flooded
583 with incoming traffic, as opposed to the permanent failure state
584 TriggerLimitIntervalSec=/TriggerLimitBurst= results in.
585 PollLimitIntervalSec=, PollLimitBurst=
587 Configures a limit on how often polling events on the file
588 descriptors backing this socket unit will be considered. This pair
589 of settings is similar to
590 TriggerLimitIntervalSec=/TriggerLimitBurst= but instead of putting
591 a (fatal) limit on the activation frequency puts a (transient)
592 limit on the polling frequency. The expected parameter syntax and
593 range are identical to that of the aforementioned options, and can
594 be disabled the same way.
595 If the polling limit is hit polling is temporarily disabled on it
597 until the specified time window passes. The polling limit hence
598 slows down connection attempts if hit, but unlike the trigger limit
599 won't cause permanent failures. It's the recommended mechanism to
600 deal with DoS attempts through packet flooding.
601 The polling limit is enforced per file descriptor to listen on, as
603 opposed to the trigger limit which is enforced for the entire
604 socket unit. This distinction matters for socket units that listen
605 on multiple file descriptors (i.e. have multiple ListenXYZ=
606 stanzas).
607 These setting defaults to 150 (in case of Accept=yes) and 15
609 (otherwise) polling events per 2s. This is considerably lower than
610 the default values for the trigger limit (see above) and means that
611 the polling limit should typically ensure the trigger limit is
612 never hit, unless one of them is reconfigured or disabled.
613 Check systemd.unit(5), systemd.exec(5), and systemd.kill(5) for more
615 settings.
616
618 systemd(1), systemctl(1), systemd-system.conf(5), systemd.unit(5),
619 systemd.exec(5), systemd.kill(5), systemd.resource-control(5),
620 systemd.service(5), systemd.directives(7), sd_listen_fds(3),
621 sd_listen_fds_with_names(3)
622 For more extensive descriptions see the "systemd for Developers"
624 series: Socket Activation[4], Socket Activation, part II[5], Converting
625 inetd Services[6], Socket Activated Internet Services and OS
626 Containers[7].
627
629 1. USB FunctionFS
630 https://docs.kernel.org/usb/functionfs.html
631 2. TCP Keepalive HOWTO
633 http://www.tldp.org/HOWTO/html_single/TCP-Keepalive-HOWTO/
634 3. Smack
636 https://docs.kernel.org/admin-guide/LSM/Smack.html
637 4. Socket Activation
639 https://0pointer.de/blog/projects/socket-activation.html
640 5. Socket Activation, part II
642 https://0pointer.de/blog/projects/socket-activation2.html
643 6. Converting inetd Services
645 https://0pointer.de/blog/projects/inetd.html
646 7. Socket Activated Internet Services and OS Containers
648 https://0pointer.de/blog/projects/socket-activated-containers.html
649systemd 254 SYSTEMD.SOCKET(5)