1 /proc/sys/net/ipv4/* Variables:
7 Forward Packets between interfaces.
9 This variable is special, its change resets all configuration
10 parameters to their default state (RFC1122 for hosts, RFC1812
13 ip_default_ttl - INTEGER
14 Default value of TTL field (Time To Live) for outgoing (but not
15 forwarded) IP packets. Should be between 1 and 255 inclusive.
16 Default: 64 (as recommended by RFC1700)
18 ip_no_pmtu_disc - INTEGER
19 Disable Path MTU Discovery. If enabled in mode 1 and a
20 fragmentation-required ICMP is received, the PMTU to this
21 destination will be set to min_pmtu (see below). You will need
22 to raise min_pmtu to the smallest interface MTU on your system
23 manually if you want to avoid locally generated fragments.
25 In mode 2 incoming Path MTU Discovery messages will be
26 discarded. Outgoing frames are handled the same as in mode 1,
27 implicitly setting IP_PMTUDISC_DONT on every created socket.
29 Mode 3 is a hardend pmtu discover mode. The kernel will only
30 accept fragmentation-needed errors if the underlying protocol
31 can verify them besides a plain socket lookup. Current
32 protocols for which pmtu events will be honored are TCP, SCTP
33 and DCCP as they verify e.g. the sequence number or the
34 association. This mode should not be enabled globally but is
35 only intended to secure e.g. name servers in namespaces where
36 TCP path mtu must still work but path MTU information of other
37 protocols should be discarded. If enabled globally this mode
38 could break other protocols.
44 default 552 - minimum discovered Path MTU
46 ip_forward_use_pmtu - BOOLEAN
47 By default we don't trust protocol path MTUs while forwarding
48 because they could be easily forged and can lead to unwanted
49 fragmentation by the router.
50 You only need to enable this if you have user-space software
51 which tries to discover path mtus by itself and depends on the
52 kernel honoring this information. This is normally not the
59 fwmark_reflect - BOOLEAN
60 Controls the fwmark of kernel-generated IPv4 reply packets that are not
61 associated with a socket for example, TCP RSTs or ICMP echo replies).
62 If unset, these packets have a fwmark of zero. If set, they have the
63 fwmark of the packet they are replying to.
66 route/max_size - INTEGER
67 Maximum number of routes allowed in the kernel. Increase
68 this when using large numbers of interfaces and/or routes.
69 From linux kernel 3.6 onwards, this is deprecated for ipv4
70 as route cache is no longer used.
72 neigh/default/gc_thresh1 - INTEGER
73 Minimum number of entries to keep. Garbage collector will not
74 purge entries if there are fewer than this number.
77 neigh/default/gc_thresh2 - INTEGER
78 Threshold when garbage collector becomes more aggressive about
79 purging entries. Entries older than 5 seconds will be cleared
80 when over this number.
83 neigh/default/gc_thresh3 - INTEGER
84 Maximum number of neighbor entries allowed. Increase this
85 when using large numbers of interfaces and when communicating
86 with large numbers of directly-connected peers.
89 neigh/default/unres_qlen_bytes - INTEGER
90 The maximum number of bytes which may be used by packets
91 queued for each unresolved address by other network layers.
93 Setting negative value is meaningless and will return error.
94 Default: 65536 Bytes(64KB)
96 neigh/default/unres_qlen - INTEGER
97 The maximum number of packets which may be queued for each
98 unresolved address by other network layers.
99 (deprecated in linux 3.3) : use unres_qlen_bytes instead.
100 Prior to linux 3.3, the default value is 3 which may cause
101 unexpected packet loss. The current default value is calculated
102 according to default value of unres_qlen_bytes and true size of
106 mtu_expires - INTEGER
107 Time, in seconds, that cached PMTU information is kept.
109 min_adv_mss - INTEGER
110 The advertised MSS depends on the first hop route MTU, but will
111 never be lower than this setting.
115 ipfrag_high_thresh - INTEGER
116 Maximum memory used to reassemble IP fragments. When
117 ipfrag_high_thresh bytes of memory is allocated for this purpose,
118 the fragment handler will toss packets until ipfrag_low_thresh
119 is reached. This also serves as a maximum limit to namespaces
120 different from the initial one.
122 ipfrag_low_thresh - INTEGER
123 Maximum memory used to reassemble IP fragments before the kernel
124 begins to remove incomplete fragment queues to free up resources.
125 The kernel still accepts new fragments for defragmentation.
127 ipfrag_time - INTEGER
128 Time in seconds to keep an IP fragment in memory.
130 ipfrag_max_dist - INTEGER
131 ipfrag_max_dist is a non-negative integer value which defines the
132 maximum "disorder" which is allowed among fragments which share a
133 common IP source address. Note that reordering of packets is
134 not unusual, but if a large number of fragments arrive from a source
135 IP address while a particular fragment queue remains incomplete, it
136 probably indicates that one or more fragments belonging to that queue
137 have been lost. When ipfrag_max_dist is positive, an additional check
138 is done on fragments before they are added to a reassembly queue - if
139 ipfrag_max_dist (or more) fragments have arrived from a particular IP
140 address between additions to any IP fragment queue using that source
141 address, it's presumed that one or more fragments in the queue are
142 lost. The existing fragment queue will be dropped, and a new one
143 started. An ipfrag_max_dist value of zero disables this check.
145 Using a very small value, e.g. 1 or 2, for ipfrag_max_dist can
146 result in unnecessarily dropping fragment queues when normal
147 reordering of packets occurs, which could lead to poor application
148 performance. Using a very large value, e.g. 50000, increases the
149 likelihood of incorrectly reassembling IP fragments that originate
150 from different IP datagrams, which could result in data corruption.
155 inet_peer_threshold - INTEGER
156 The approximate size of the storage. Starting from this threshold
157 entries will be thrown aggressively. This threshold also determines
158 entries' time-to-live and time intervals between garbage collection
159 passes. More entries, less time-to-live, less GC interval.
161 inet_peer_minttl - INTEGER
162 Minimum time-to-live of entries. Should be enough to cover fragment
163 time-to-live on the reassembling side. This minimum time-to-live is
164 guaranteed if the pool size is less than inet_peer_threshold.
167 inet_peer_maxttl - INTEGER
168 Maximum time-to-live of entries. Unused entries will expire after
169 this period of time if there is no memory pressure on the pool (i.e.
170 when the number of entries in the pool is very small).
176 Limit of socket listen() backlog, known in userspace as SOMAXCONN.
177 Defaults to 128. See also tcp_max_syn_backlog for additional tuning
180 tcp_abort_on_overflow - BOOLEAN
181 If listening service is too slow to accept new connections,
182 reset them. Default state is FALSE. It means that if overflow
183 occurred due to a burst, connection will recover. Enable this
184 option _only_ if you are really sure that listening daemon
185 cannot be tuned to accept connections faster. Enabling this
186 option can harm clients of your server.
188 tcp_adv_win_scale - INTEGER
189 Count buffering overhead as bytes/2^tcp_adv_win_scale
190 (if tcp_adv_win_scale > 0) or bytes-bytes/2^(-tcp_adv_win_scale),
192 Possible values are [-31, 31], inclusive.
195 tcp_allowed_congestion_control - STRING
196 Show/set the congestion control choices available to non-privileged
197 processes. The list is a subset of those listed in
198 tcp_available_congestion_control.
199 Default is "reno" and the default setting (tcp_congestion_control).
201 tcp_app_win - INTEGER
202 Reserve max(window/2^tcp_app_win, mss) of window for application
203 buffer. Value 0 is special, it means that nothing is reserved.
206 tcp_autocorking - BOOLEAN
207 Enable TCP auto corking :
208 When applications do consecutive small write()/sendmsg() system calls,
209 we try to coalesce these small writes as much as possible, to lower
210 total amount of sent packets. This is done if at least one prior
211 packet for the flow is waiting in Qdisc queues or device transmit
212 queue. Applications can still use TCP_CORK for optimal behavior
213 when they know how/when to uncork their sockets.
216 tcp_available_congestion_control - STRING
217 Shows the available congestion control choices that are registered.
218 More congestion control algorithms may be available as modules,
221 tcp_base_mss - INTEGER
222 The initial value of search_low to be used by the packetization layer
223 Path MTU discovery (MTU probing). If MTU probing is enabled,
224 this is the initial MSS used by the connection.
226 tcp_congestion_control - STRING
227 Set the congestion control algorithm to be used for new
228 connections. The algorithm "reno" is always available, but
229 additional choices may be available based on kernel configuration.
230 Default is set as part of kernel configuration.
231 For passive connections, the listener congestion control choice
233 [see setsockopt(listenfd, SOL_TCP, TCP_CONGESTION, "name" ...) ]
236 Allows TCP to send "duplicate" SACKs.
238 tcp_early_retrans - INTEGER
239 Enable Early Retransmit (ER), per RFC 5827. ER lowers the threshold
240 for triggering fast retransmit when the amount of outstanding data is
241 small and when no previously unsent data can be transmitted (such
242 that limited transmit could be used). Also controls the use of
243 Tail loss probe (TLP) that converts RTOs occurring due to tail
244 losses into fast recovery (draft-dukkipati-tcpm-tcp-loss-probe-01).
248 2 enables ER but delays fast recovery and fast retransmit
249 by a fourth of RTT. This mitigates connection falsely
250 recovers when network has a small degree of reordering
251 (less than 3 packets).
252 3 enables delayed ER and TLP.
257 Control use of Explicit Congestion Notification (ECN) by TCP.
258 ECN is used only when both ends of the TCP connection indicate
259 support for it. This feature is useful in avoiding losses due
260 to congestion by allowing supporting routers to signal
261 congestion before having to drop packets.
263 0 Disable ECN. Neither initiate nor accept ECN.
264 1 Enable ECN when requested by incoming connections and
265 also request ECN on outgoing connection attempts.
266 2 Enable ECN when requested by incoming connections
267 but do not request ECN on outgoing connections.
270 tcp_ecn_fallback - BOOLEAN
271 If the kernel detects that ECN connection misbehaves, enable fall
272 back to non-ECN. Currently, this knob implements the fallback
273 from RFC3168, section 6.1.1.1., but we reserve that in future,
274 additional detection mechanisms could be implemented under this
275 knob. The value is not used, if tcp_ecn or per route (or congestion
276 control) ECN settings are disabled.
277 Default: 1 (fallback enabled)
280 Enable FACK congestion avoidance and fast retransmission.
281 The value is not used, if tcp_sack is not enabled.
283 tcp_fin_timeout - INTEGER
284 The length of time an orphaned (no longer referenced by any
285 application) connection will remain in the FIN_WAIT_2 state
286 before it is aborted at the local end. While a perfectly
287 valid "receive only" state for an un-orphaned connection, an
288 orphaned connection in FIN_WAIT_2 state could otherwise wait
289 forever for the remote to close its end of the connection.
294 Enables Forward RTO-Recovery (F-RTO) defined in RFC5682.
295 F-RTO is an enhanced recovery algorithm for TCP retransmission
296 timeouts. It is particularly beneficial in networks where the
297 RTT fluctuates (e.g., wireless). F-RTO is sender-side only
298 modification. It does not require any support from the peer.
300 By default it's enabled with a non-zero value. 0 disables F-RTO.
302 tcp_invalid_ratelimit - INTEGER
303 Limit the maximal rate for sending duplicate acknowledgments
304 in response to incoming TCP packets that are for an existing
305 connection but that are invalid due to any of these reasons:
307 (a) out-of-window sequence number,
308 (b) out-of-window acknowledgment number, or
309 (c) PAWS (Protection Against Wrapped Sequence numbers) check failure
311 This can help mitigate simple "ack loop" DoS attacks, wherein
312 a buggy or malicious middlebox or man-in-the-middle can
313 rewrite TCP header fields in manner that causes each endpoint
314 to think that the other is sending invalid TCP segments, thus
315 causing each side to send an unterminating stream of duplicate
316 acknowledgments for invalid segments.
318 Using 0 disables rate-limiting of dupacks in response to
319 invalid segments; otherwise this value specifies the minimal
320 space between sending such dupacks, in milliseconds.
322 Default: 500 (milliseconds).
324 tcp_keepalive_time - INTEGER
325 How often TCP sends out keepalive messages when keepalive is enabled.
328 tcp_keepalive_probes - INTEGER
329 How many keepalive probes TCP sends out, until it decides that the
330 connection is broken. Default value: 9.
332 tcp_keepalive_intvl - INTEGER
333 How frequently the probes are send out. Multiplied by
334 tcp_keepalive_probes it is time to kill not responding connection,
335 after probes started. Default value: 75sec i.e. connection
336 will be aborted after ~11 minutes of retries.
338 tcp_low_latency - BOOLEAN
339 If set, the TCP stack makes decisions that prefer lower
340 latency as opposed to higher throughput. By default, this
341 option is not set meaning that higher throughput is preferred.
342 An example of an application where this default should be
343 changed would be a Beowulf compute cluster.
346 tcp_max_orphans - INTEGER
347 Maximal number of TCP sockets not attached to any user file handle,
348 held by system. If this number is exceeded orphaned connections are
349 reset immediately and warning is printed. This limit exists
350 only to prevent simple DoS attacks, you _must_ not rely on this
351 or lower the limit artificially, but rather increase it
352 (probably, after increasing installed memory),
353 if network conditions require more than default value,
354 and tune network services to linger and kill such states
355 more aggressively. Let me to remind again: each orphan eats
356 up to ~64K of unswappable memory.
358 tcp_max_syn_backlog - INTEGER
359 Maximal number of remembered connection requests, which have not
360 received an acknowledgment from connecting client.
361 The minimal value is 128 for low memory machines, and it will
362 increase in proportion to the memory of machine.
363 If server suffers from overload, try increasing this number.
365 tcp_max_tw_buckets - INTEGER
366 Maximal number of timewait sockets held by system simultaneously.
367 If this number is exceeded time-wait socket is immediately destroyed
368 and warning is printed. This limit exists only to prevent
369 simple DoS attacks, you _must_ not lower the limit artificially,
370 but rather increase it (probably, after increasing installed memory),
371 if network conditions require more than default value.
373 tcp_mem - vector of 3 INTEGERs: min, pressure, max
374 min: below this number of pages TCP is not bothered about its
377 pressure: when amount of memory allocated by TCP exceeds this number
378 of pages, TCP moderates its memory consumption and enters memory
379 pressure mode, which is exited when memory consumption falls
382 max: number of pages allowed for queueing by all TCP sockets.
384 Defaults are calculated at boot time from amount of available
387 tcp_min_rtt_wlen - INTEGER
388 The window length of the windowed min filter to track the minimum RTT.
389 A shorter window lets a flow more quickly pick up new (higher)
390 minimum RTT when it is moved to a longer path (e.g., due to traffic
391 engineering). A longer window makes the filter more resistant to RTT
392 inflations such as transient congestion. The unit is seconds.
395 tcp_moderate_rcvbuf - BOOLEAN
396 If set, TCP performs receive buffer auto-tuning, attempting to
397 automatically size the buffer (no greater than tcp_rmem[2]) to
398 match the size required by the path for full throughput. Enabled by
401 tcp_mtu_probing - INTEGER
402 Controls TCP Packetization-Layer Path MTU Discovery. Takes three
405 1 - Disabled by default, enabled when an ICMP black hole detected
406 2 - Always enabled, use initial MSS of tcp_base_mss.
408 tcp_probe_interval - INTEGER
409 Controls how often to start TCP Packetization-Layer Path MTU
410 Discovery reprobe. The default is reprobing every 10 minutes as
413 tcp_probe_threshold - INTEGER
414 Controls when TCP Packetization-Layer Path MTU Discovery probing
415 will stop in respect to the width of search range in bytes. Default
418 tcp_no_metrics_save - BOOLEAN
419 By default, TCP saves various connection metrics in the route cache
420 when the connection closes, so that connections established in the
421 near future can use these to set initial conditions. Usually, this
422 increases overall performance, but may sometimes cause performance
423 degradation. If set, TCP will not cache metrics on closing
426 tcp_orphan_retries - INTEGER
427 This value influences the timeout of a locally closed TCP connection,
428 when RTO retransmissions remain unacknowledged.
429 See tcp_retries2 for more details.
431 The default value is 8.
432 If your machine is a loaded WEB server,
433 you should think about lowering this value, such sockets
434 may consume significant resources. Cf. tcp_max_orphans.
436 tcp_reordering - INTEGER
437 Initial reordering level of packets in a TCP stream.
438 TCP stack can then dynamically adjust flow reordering level
439 between this initial value and tcp_max_reordering
442 tcp_max_reordering - INTEGER
443 Maximal reordering level of packets in a TCP stream.
444 300 is a fairly conservative value, but you might increase it
445 if paths are using per packet load balancing (like bonding rr mode)
448 tcp_retrans_collapse - BOOLEAN
449 Bug-to-bug compatibility with some broken printers.
450 On retransmit try to send bigger packets to work around bugs in
453 tcp_retries1 - INTEGER
454 This value influences the time, after which TCP decides, that
455 something is wrong due to unacknowledged RTO retransmissions,
456 and reports this suspicion to the network layer.
457 See tcp_retries2 for more details.
459 RFC 1122 recommends at least 3 retransmissions, which is the
462 tcp_retries2 - INTEGER
463 This value influences the timeout of an alive TCP connection,
464 when RTO retransmissions remain unacknowledged.
465 Given a value of N, a hypothetical TCP connection following
466 exponential backoff with an initial RTO of TCP_RTO_MIN would
467 retransmit N times before killing the connection at the (N+1)th RTO.
469 The default value of 15 yields a hypothetical timeout of 924.6
470 seconds and is a lower bound for the effective timeout.
471 TCP will effectively time out at the first RTO which exceeds the
472 hypothetical timeout.
474 RFC 1122 recommends at least 100 seconds for the timeout,
475 which corresponds to a value of at least 8.
477 tcp_rfc1337 - BOOLEAN
478 If set, the TCP stack behaves conforming to RFC1337. If unset,
479 we are not conforming to RFC, but prevent TCP TIME_WAIT
483 tcp_rmem - vector of 3 INTEGERs: min, default, max
484 min: Minimal size of receive buffer used by TCP sockets.
485 It is guaranteed to each TCP socket, even under moderate memory
489 default: initial size of receive buffer used by TCP sockets.
490 This value overrides net.core.rmem_default used by other protocols.
491 Default: 87380 bytes. This value results in window of 65535 with
492 default setting of tcp_adv_win_scale and tcp_app_win:0 and a bit
493 less for default tcp_app_win. See below about these variables.
495 max: maximal size of receive buffer allowed for automatically
496 selected receiver buffers for TCP socket. This value does not override
497 net.core.rmem_max. Calling setsockopt() with SO_RCVBUF disables
498 automatic tuning of that socket's receive buffer size, in which
499 case this value is ignored.
500 Default: between 87380B and 6MB, depending on RAM size.
503 Enable select acknowledgments (SACKS).
505 tcp_slow_start_after_idle - BOOLEAN
506 If set, provide RFC2861 behavior and time out the congestion
507 window after an idle period. An idle period is defined at
508 the current RTO. If unset, the congestion window will not
509 be timed out after an idle period.
513 Use the Host requirements interpretation of the TCP urgent pointer field.
514 Most hosts use the older BSD interpretation, so if you turn this on
515 Linux might not communicate correctly with them.
518 tcp_synack_retries - INTEGER
519 Number of times SYNACKs for a passive TCP connection attempt will
520 be retransmitted. Should not be higher than 255. Default value
521 is 5, which corresponds to 31seconds till the last retransmission
522 with the current initial RTO of 1second. With this the final timeout
523 for a passive TCP connection will happen after 63seconds.
525 tcp_syncookies - BOOLEAN
526 Only valid when the kernel was compiled with CONFIG_SYN_COOKIES
527 Send out syncookies when the syn backlog queue of a socket
528 overflows. This is to prevent against the common 'SYN flood attack'
531 Note, that syncookies is fallback facility.
532 It MUST NOT be used to help highly loaded servers to stand
533 against legal connection rate. If you see SYN flood warnings
534 in your logs, but investigation shows that they occur
535 because of overload with legal connections, you should tune
536 another parameters until this warning disappear.
537 See: tcp_max_syn_backlog, tcp_synack_retries, tcp_abort_on_overflow.
539 syncookies seriously violate TCP protocol, do not allow
540 to use TCP extensions, can result in serious degradation
541 of some services (f.e. SMTP relaying), visible not by you,
542 but your clients and relays, contacting you. While you see
543 SYN flood warnings in logs not being really flooded, your server
544 is seriously misconfigured.
546 If you want to test which effects syncookies have to your
547 network connections you can set this knob to 2 to enable
548 unconditionally generation of syncookies.
550 tcp_fastopen - INTEGER
551 Enable TCP Fast Open feature (draft-ietf-tcpm-fastopen) to send data
552 in the opening SYN packet. To use this feature, the client application
553 must use sendmsg() or sendto() with MSG_FASTOPEN flag rather than
554 connect() to perform a TCP handshake automatically.
556 The values (bitmap) are
557 1: Enables sending data in the opening SYN on the client w/ MSG_FASTOPEN.
558 2: Enables TCP Fast Open on the server side, i.e., allowing data in
559 a SYN packet to be accepted and passed to the application before
560 3-way hand shake finishes.
561 4: Send data in the opening SYN regardless of cookie availability and
562 without a cookie option.
563 0x100: Accept SYN data w/o validating the cookie.
564 0x200: Accept data-in-SYN w/o any cookie option present.
565 0x400/0x800: Enable Fast Open on all listeners regardless of the
566 TCP_FASTOPEN socket option. The two different flags designate two
567 different ways of setting max_qlen without the TCP_FASTOPEN socket
572 Note that the client & server side Fast Open flags (1 and 2
573 respectively) must be also enabled before the rest of flags can take
576 See include/net/tcp.h and the code for more details.
578 tcp_syn_retries - INTEGER
579 Number of times initial SYNs for an active TCP connection attempt
580 will be retransmitted. Should not be higher than 255. Default value
581 is 6, which corresponds to 63seconds till the last retransmission
582 with the current initial RTO of 1second. With this the final timeout
583 for an active TCP connection attempt will happen after 127seconds.
585 tcp_timestamps - BOOLEAN
586 Enable timestamps as defined in RFC1323.
588 tcp_min_tso_segs - INTEGER
589 Minimal number of segments per TSO frame.
590 Since linux-3.12, TCP does an automatic sizing of TSO frames,
591 depending on flow rate, instead of filling 64Kbytes packets.
592 For specific usages, it's possible to force TCP to build big
593 TSO frames. Note that TCP stack might split too big TSO packets
594 if available window is too small.
597 tcp_pacing_ss_ratio - INTEGER
598 sk->sk_pacing_rate is set by TCP stack using a ratio applied
599 to current rate. (current_rate = cwnd * mss / srtt)
600 If TCP is in slow start, tcp_pacing_ss_ratio is applied
601 to let TCP probe for bigger speeds, assuming cwnd can be
602 doubled every other RTT.
605 tcp_pacing_ca_ratio - INTEGER
606 sk->sk_pacing_rate is set by TCP stack using a ratio applied
607 to current rate. (current_rate = cwnd * mss / srtt)
608 If TCP is in congestion avoidance phase, tcp_pacing_ca_ratio
609 is applied to conservatively probe for bigger throughput.
612 tcp_tso_win_divisor - INTEGER
613 This allows control over what percentage of the congestion window
614 can be consumed by a single TSO frame.
615 The setting of this parameter is a choice between burstiness and
616 building larger TSO frames.
619 tcp_tw_recycle - BOOLEAN
620 Enable fast recycling TIME-WAIT sockets. Default value is 0.
621 It should not be changed without advice/request of technical
624 tcp_tw_reuse - BOOLEAN
625 Allow to reuse TIME-WAIT sockets for new connections when it is
626 safe from protocol viewpoint. Default value is 0.
627 It should not be changed without advice/request of technical
630 tcp_window_scaling - BOOLEAN
631 Enable window scaling as defined in RFC1323.
633 tcp_wmem - vector of 3 INTEGERs: min, default, max
634 min: Amount of memory reserved for send buffers for TCP sockets.
635 Each TCP socket has rights to use it due to fact of its birth.
638 default: initial size of send buffer used by TCP sockets. This
639 value overrides net.core.wmem_default used by other protocols.
640 It is usually lower than net.core.wmem_default.
643 max: Maximal amount of memory allowed for automatically tuned
644 send buffers for TCP sockets. This value does not override
645 net.core.wmem_max. Calling setsockopt() with SO_SNDBUF disables
646 automatic tuning of that socket's send buffer size, in which case
647 this value is ignored.
648 Default: between 64K and 4MB, depending on RAM size.
650 tcp_notsent_lowat - UNSIGNED INTEGER
651 A TCP socket can control the amount of unsent bytes in its write queue,
652 thanks to TCP_NOTSENT_LOWAT socket option. poll()/select()/epoll()
653 reports POLLOUT events if the amount of unsent bytes is below a per
654 socket value, and if the write queue is not full. sendmsg() will
655 also not add new buffers if the limit is hit.
657 This global variable controls the amount of unsent data for
658 sockets not using TCP_NOTSENT_LOWAT. For these sockets, a change
659 to the global variable has immediate effect.
661 Default: UINT_MAX (0xFFFFFFFF)
663 tcp_workaround_signed_windows - BOOLEAN
664 If set, assume no receipt of a window scaling option means the
665 remote TCP is broken and treats the window as a signed quantity.
666 If unset, assume the remote TCP is not broken even if we do
667 not receive a window scaling option from them.
670 tcp_thin_linear_timeouts - BOOLEAN
671 Enable dynamic triggering of linear timeouts for thin streams.
672 If set, a check is performed upon retransmission by timeout to
673 determine if the stream is thin (less than 4 packets in flight).
674 As long as the stream is found to be thin, up to 6 linear
675 timeouts may be performed before exponential backoff mode is
676 initiated. This improves retransmission latency for
677 non-aggressive thin streams, often found to be time-dependent.
678 For more information on thin streams, see
679 Documentation/networking/tcp-thin.txt
682 tcp_thin_dupack - BOOLEAN
683 Enable dynamic triggering of retransmissions after one dupACK
684 for thin streams. If set, a check is performed upon reception
685 of a dupACK to determine if the stream is thin (less than 4
686 packets in flight). As long as the stream is found to be thin,
687 data is retransmitted on the first received dupACK. This
688 improves retransmission latency for non-aggressive thin
689 streams, often found to be time-dependent.
690 For more information on thin streams, see
691 Documentation/networking/tcp-thin.txt
694 tcp_limit_output_bytes - INTEGER
695 Controls TCP Small Queue limit per tcp socket.
696 TCP bulk sender tends to increase packets in flight until it
697 gets losses notifications. With SNDBUF autotuning, this can
698 result in a large amount of packets queued in qdisc/device
699 on the local machine, hurting latency of other flows, for
700 typical pfifo_fast qdiscs.
701 tcp_limit_output_bytes limits the number of bytes on qdisc
702 or device to reduce artificial RTT/cwnd and reduce bufferbloat.
705 tcp_challenge_ack_limit - INTEGER
706 Limits number of Challenge ACK sent per second, as recommended
707 in RFC 5961 (Improving TCP's Robustness to Blind In-Window Attacks)
712 udp_mem - vector of 3 INTEGERs: min, pressure, max
713 Number of pages allowed for queueing by all UDP sockets.
715 min: Below this number of pages UDP is not bothered about its
716 memory appetite. When amount of memory allocated by UDP exceeds
717 this number, UDP starts to moderate memory usage.
719 pressure: This value was introduced to follow format of tcp_mem.
721 max: Number of pages allowed for queueing by all UDP sockets.
723 Default is calculated at boot time from amount of available memory.
725 udp_rmem_min - INTEGER
726 Minimal size of receive buffer used by UDP sockets in moderation.
727 Each UDP socket is able to use the size for receiving data, even if
728 total pages of UDP sockets exceed udp_mem pressure. The unit is byte.
731 udp_wmem_min - INTEGER
732 Minimal size of send buffer used by UDP sockets in moderation.
733 Each UDP socket is able to use the size for sending data, even if
734 total pages of UDP sockets exceed udp_mem pressure. The unit is byte.
739 cipso_cache_enable - BOOLEAN
740 If set, enable additions to and lookups from the CIPSO label mapping
741 cache. If unset, additions are ignored and lookups always result in a
742 miss. However, regardless of the setting the cache is still
743 invalidated when required when means you can safely toggle this on and
744 off and the cache will always be "safe".
747 cipso_cache_bucket_size - INTEGER
748 The CIPSO label cache consists of a fixed size hash table with each
749 hash bucket containing a number of cache entries. This variable limits
750 the number of entries in each hash bucket; the larger the value the
751 more CIPSO label mappings that can be cached. When the number of
752 entries in a given hash bucket reaches this limit adding new entries
753 causes the oldest entry in the bucket to be removed to make room.
756 cipso_rbm_optfmt - BOOLEAN
757 Enable the "Optimized Tag 1 Format" as defined in section 3.4.2.6 of
758 the CIPSO draft specification (see Documentation/netlabel for details).
759 This means that when set the CIPSO tag will be padded with empty
760 categories in order to make the packet data 32-bit aligned.
763 cipso_rbm_structvalid - BOOLEAN
764 If set, do a very strict check of the CIPSO option when
765 ip_options_compile() is called. If unset, relax the checks done during
766 ip_options_compile(). Either way is "safe" as errors are caught else
767 where in the CIPSO processing code but setting this to 0 (False) should
768 result in less work (i.e. it should be faster) but could cause problems
769 with other implementations that require strict checking.
774 ip_local_port_range - 2 INTEGERS
775 Defines the local port range that is used by TCP and UDP to
776 choose the local port. The first number is the first, the
777 second the last local port number.
778 If possible, it is better these numbers have different parity.
779 (one even and one odd values)
780 The default values are 32768 and 60999 respectively.
782 ip_local_reserved_ports - list of comma separated ranges
783 Specify the ports which are reserved for known third-party
784 applications. These ports will not be used by automatic port
785 assignments (e.g. when calling connect() or bind() with port
786 number 0). Explicit port allocation behavior is unchanged.
788 The format used for both input and output is a comma separated
789 list of ranges (e.g. "1,2-4,10-10" for ports 1, 2, 3, 4 and
790 10). Writing to the file will clear all previously reserved
791 ports and update the current list with the one given in the
794 Note that ip_local_port_range and ip_local_reserved_ports
795 settings are independent and both are considered by the kernel
796 when determining which ports are available for automatic port
799 You can reserve ports which are not in the current
800 ip_local_port_range, e.g.:
802 $ cat /proc/sys/net/ipv4/ip_local_port_range
804 $ cat /proc/sys/net/ipv4/ip_local_reserved_ports
807 although this is redundant. However such a setting is useful
808 if later the port range is changed to a value that will
809 include the reserved ports.
813 ip_nonlocal_bind - BOOLEAN
814 If set, allows processes to bind() to non-local IP addresses,
815 which can be quite useful - but may break some applications.
819 If set non-zero, enables support for dynamic addresses.
820 If set to a non-zero value larger than 1, a kernel log
821 message will be printed when dynamic address rewriting
825 ip_early_demux - BOOLEAN
826 Optimize input packet processing down to one demux for
827 certain kinds of local sockets. Currently we only do this
828 for established TCP sockets.
830 It may add an additional cost for pure routing workloads that
831 reduces overall throughput, in such case you should disable it.
834 icmp_echo_ignore_all - BOOLEAN
835 If set non-zero, then the kernel will ignore all ICMP ECHO
839 icmp_echo_ignore_broadcasts - BOOLEAN
840 If set non-zero, then the kernel will ignore all ICMP ECHO and
841 TIMESTAMP requests sent to it via broadcast/multicast.
844 icmp_ratelimit - INTEGER
845 Limit the maximal rates for sending ICMP packets whose type matches
846 icmp_ratemask (see below) to specific targets.
847 0 to disable any limiting,
848 otherwise the minimal space between responses in milliseconds.
849 Note that another sysctl, icmp_msgs_per_sec limits the number
850 of ICMP packets sent on all targets.
853 icmp_msgs_per_sec - INTEGER
854 Limit maximal number of ICMP packets sent per second from this host.
855 Only messages whose type matches icmp_ratemask (see below) are
856 controlled by this limit.
859 icmp_msgs_burst - INTEGER
860 icmp_msgs_per_sec controls number of ICMP packets sent per second,
861 while icmp_msgs_burst controls the burst size of these packets.
864 icmp_ratemask - INTEGER
865 Mask made of ICMP types for which rates are being limited.
866 Significant bits: IHGFEDCBA9876543210
867 Default mask: 0000001100000011000 (6168)
869 Bit definitions (see include/linux/icmp.h):
871 3 Destination Unreachable *
876 C Parameter Problem *
881 H Address Mask Request
884 * These are rate limited by default (see default mask above)
886 icmp_ignore_bogus_error_responses - BOOLEAN
887 Some routers violate RFC1122 by sending bogus responses to broadcast
888 frames. Such violations are normally logged via a kernel warning.
889 If this is set to TRUE, the kernel will not give such warnings, which
890 will avoid log file clutter.
893 icmp_errors_use_inbound_ifaddr - BOOLEAN
895 If zero, icmp error messages are sent with the primary address of
896 the exiting interface.
898 If non-zero, the message will be sent with the primary address of
899 the interface that received the packet that caused the icmp error.
900 This is the behaviour network many administrators will expect from
901 a router. And it can make debugging complicated network layouts
904 Note that if no primary address exists for the interface selected,
905 then the primary address of the first non-loopback interface that
906 has one will be used regardless of this setting.
910 igmp_max_memberships - INTEGER
911 Change the maximum number of multicast groups we can subscribe to.
914 Theoretical maximum value is bounded by having to send a membership
915 report in a single datagram (i.e. the report can't span multiple
916 datagrams, or risk confusing the switch and leaving groups you don't
919 The number of supported groups 'M' is bounded by the number of group
920 report entries you can fit into a single datagram of 65535 bytes.
922 M = 65536-sizeof (ip header)/(sizeof(Group record))
924 Group records are variable length, with a minimum of 12 bytes.
925 So net.ipv4.igmp_max_memberships should not be set higher than:
927 (65536-24) / 12 = 5459
929 The value 5459 assumes no IP header options, so in practice
930 this number may be lower.
932 conf/interface/* changes special settings per interface (where
933 "interface" is the name of your network interface)
935 conf/all/* is special, changes the settings for all interfaces
938 Controls the IGMP query robustness variable (see RFC2236 8.1).
939 Default: 2 (as specified by RFC2236 8.1)
940 Minimum: 1 (as specified by RFC6636 4.5)
942 log_martians - BOOLEAN
943 Log packets with impossible addresses to kernel log.
944 log_martians for the interface will be enabled if at least one of
945 conf/{all,interface}/log_martians is set to TRUE,
946 it will be disabled otherwise
948 accept_redirects - BOOLEAN
949 Accept ICMP redirect messages.
950 accept_redirects for the interface will be enabled if:
951 - both conf/{all,interface}/accept_redirects are TRUE in the case
952 forwarding for the interface is enabled
954 - at least one of conf/{all,interface}/accept_redirects is TRUE in the
955 case forwarding for the interface is disabled
956 accept_redirects for the interface will be disabled otherwise
961 Enable IP forwarding on this interface.
963 mc_forwarding - BOOLEAN
964 Do multicast routing. The kernel needs to be compiled with CONFIG_MROUTE
965 and a multicast routing daemon is required.
966 conf/all/mc_forwarding must also be set to TRUE to enable multicast
967 routing for the interface
970 Integer value used to differentiate the devices by the medium they
971 are attached to. Two devices can have different id values when
972 the broadcast packets are received only on one of them.
973 The default value 0 means that the device is the only interface
974 to its medium, value of -1 means that medium is not known.
976 Currently, it is used to change the proxy_arp behavior:
977 the proxy_arp feature is enabled for packets forwarded between
978 two devices attached to different media.
982 proxy_arp for the interface will be enabled if at least one of
983 conf/{all,interface}/proxy_arp is set to TRUE,
984 it will be disabled otherwise
986 proxy_arp_pvlan - BOOLEAN
987 Private VLAN proxy arp.
988 Basically allow proxy arp replies back to the same interface
989 (from which the ARP request/solicitation was received).
991 This is done to support (ethernet) switch features, like RFC
992 3069, where the individual ports are NOT allowed to
993 communicate with each other, but they are allowed to talk to
994 the upstream router. As described in RFC 3069, it is possible
995 to allow these hosts to communicate through the upstream
996 router by proxy_arp'ing. Don't need to be used together with
999 This technology is known by different names:
1000 In RFC 3069 it is called VLAN Aggregation.
1001 Cisco and Allied Telesyn call it Private VLAN.
1002 Hewlett-Packard call it Source-Port filtering or port-isolation.
1003 Ericsson call it MAC-Forced Forwarding (RFC Draft).
1005 shared_media - BOOLEAN
1006 Send(router) or accept(host) RFC1620 shared media redirects.
1007 Overrides ip_secure_redirects.
1008 shared_media for the interface will be enabled if at least one of
1009 conf/{all,interface}/shared_media is set to TRUE,
1010 it will be disabled otherwise
1013 secure_redirects - BOOLEAN
1014 Accept ICMP redirect messages only for gateways,
1015 listed in default gateway list.
1016 secure_redirects for the interface will be enabled if at least one of
1017 conf/{all,interface}/secure_redirects is set to TRUE,
1018 it will be disabled otherwise
1021 send_redirects - BOOLEAN
1022 Send redirects, if router.
1023 send_redirects for the interface will be enabled if at least one of
1024 conf/{all,interface}/send_redirects is set to TRUE,
1025 it will be disabled otherwise
1028 bootp_relay - BOOLEAN
1029 Accept packets with source address 0.b.c.d destined
1030 not to this host as local ones. It is supposed, that
1031 BOOTP relay daemon will catch and forward such packets.
1032 conf/all/bootp_relay must also be set to TRUE to enable BOOTP relay
1035 Not Implemented Yet.
1037 accept_source_route - BOOLEAN
1038 Accept packets with SRR option.
1039 conf/all/accept_source_route must also be set to TRUE to accept packets
1040 with SRR option on the interface
1041 default TRUE (router)
1044 accept_local - BOOLEAN
1045 Accept packets with local source addresses. In combination with
1046 suitable routing, this can be used to direct packets between two
1047 local interfaces over the wire and have them accepted properly.
1050 route_localnet - BOOLEAN
1051 Do not consider loopback addresses as martian source or destination
1052 while routing. This enables the use of 127/8 for local routing purposes.
1056 0 - No source validation.
1057 1 - Strict mode as defined in RFC3704 Strict Reverse Path
1058 Each incoming packet is tested against the FIB and if the interface
1059 is not the best reverse path the packet check will fail.
1060 By default failed packets are discarded.
1061 2 - Loose mode as defined in RFC3704 Loose Reverse Path
1062 Each incoming packet's source address is also tested against the FIB
1063 and if the source address is not reachable via any interface
1064 the packet check will fail.
1066 Current recommended practice in RFC3704 is to enable strict mode
1067 to prevent IP spoofing from DDos attacks. If using asymmetric routing
1068 or other complicated routing, then loose mode is recommended.
1070 The max value from conf/{all,interface}/rp_filter is used
1071 when doing source validation on the {interface}.
1073 Default value is 0. Note that some distributions enable it
1076 arp_filter - BOOLEAN
1077 1 - Allows you to have multiple network interfaces on the same
1078 subnet, and have the ARPs for each interface be answered
1079 based on whether or not the kernel would route a packet from
1080 the ARP'd IP out that interface (therefore you must use source
1081 based routing for this to work). In other words it allows control
1082 of which cards (usually 1) will respond to an arp request.
1084 0 - (default) The kernel can respond to arp requests with addresses
1085 from other interfaces. This may seem wrong but it usually makes
1086 sense, because it increases the chance of successful communication.
1087 IP addresses are owned by the complete host on Linux, not by
1088 particular interfaces. Only for more complex setups like load-
1089 balancing, does this behaviour cause problems.
1091 arp_filter for the interface will be enabled if at least one of
1092 conf/{all,interface}/arp_filter is set to TRUE,
1093 it will be disabled otherwise
1095 arp_announce - INTEGER
1096 Define different restriction levels for announcing the local
1097 source IP address from IP packets in ARP requests sent on
1099 0 - (default) Use any local address, configured on any interface
1100 1 - Try to avoid local addresses that are not in the target's
1101 subnet for this interface. This mode is useful when target
1102 hosts reachable via this interface require the source IP
1103 address in ARP requests to be part of their logical network
1104 configured on the receiving interface. When we generate the
1105 request we will check all our subnets that include the
1106 target IP and will preserve the source address if it is from
1107 such subnet. If there is no such subnet we select source
1108 address according to the rules for level 2.
1109 2 - Always use the best local address for this target.
1110 In this mode we ignore the source address in the IP packet
1111 and try to select local address that we prefer for talks with
1112 the target host. Such local address is selected by looking
1113 for primary IP addresses on all our subnets on the outgoing
1114 interface that include the target IP address. If no suitable
1115 local address is found we select the first local address
1116 we have on the outgoing interface or on all other interfaces,
1117 with the hope we will receive reply for our request and
1118 even sometimes no matter the source IP address we announce.
1120 The max value from conf/{all,interface}/arp_announce is used.
1122 Increasing the restriction level gives more chance for
1123 receiving answer from the resolved target while decreasing
1124 the level announces more valid sender's information.
1126 arp_ignore - INTEGER
1127 Define different modes for sending replies in response to
1128 received ARP requests that resolve local target IP addresses:
1129 0 - (default): reply for any local target IP address, configured
1131 1 - reply only if the target IP address is local address
1132 configured on the incoming interface
1133 2 - reply only if the target IP address is local address
1134 configured on the incoming interface and both with the
1135 sender's IP address are part from same subnet on this interface
1136 3 - do not reply for local addresses configured with scope host,
1137 only resolutions for global and link addresses are replied
1139 8 - do not reply for all local addresses
1141 The max value from conf/{all,interface}/arp_ignore is used
1142 when ARP request is received on the {interface}
1144 arp_notify - BOOLEAN
1145 Define mode for notification of address and device changes.
1146 0 - (default): do nothing
1147 1 - Generate gratuitous arp requests when device is brought up
1148 or hardware address changes.
1150 arp_accept - BOOLEAN
1151 Define behavior for gratuitous ARP frames who's IP is not
1152 already present in the ARP table:
1153 0 - don't create new entries in the ARP table
1154 1 - create new entries in the ARP table
1156 Both replies and requests type gratuitous arp will trigger the
1157 ARP table to be updated, if this setting is on.
1159 If the ARP table already contains the IP address of the
1160 gratuitous arp frame, the arp table will be updated regardless
1161 if this setting is on or off.
1163 mcast_solicit - INTEGER
1164 The maximum number of multicast probes in INCOMPLETE state,
1165 when the associated hardware address is unknown. Defaults
1168 ucast_solicit - INTEGER
1169 The maximum number of unicast probes in PROBE state, when
1170 the hardware address is being reconfirmed. Defaults to 3.
1172 app_solicit - INTEGER
1173 The maximum number of probes to send to the user space ARP daemon
1174 via netlink before dropping back to multicast probes (see
1175 mcast_resolicit). Defaults to 0.
1177 mcast_resolicit - INTEGER
1178 The maximum number of multicast probes after unicast and
1179 app probes in PROBE state. Defaults to 0.
1181 disable_policy - BOOLEAN
1182 Disable IPSEC policy (SPD) for this interface
1184 disable_xfrm - BOOLEAN
1185 Disable IPSEC encryption on this interface, whatever the policy
1187 igmpv2_unsolicited_report_interval - INTEGER
1188 The interval in milliseconds in which the next unsolicited
1189 IGMPv1 or IGMPv2 report retransmit will take place.
1190 Default: 10000 (10 seconds)
1192 igmpv3_unsolicited_report_interval - INTEGER
1193 The interval in milliseconds in which the next unsolicited
1194 IGMPv3 report retransmit will take place.
1195 Default: 1000 (1 seconds)
1197 promote_secondaries - BOOLEAN
1198 When a primary IP address is removed from this interface
1199 promote a corresponding secondary IP address instead of
1200 removing all the corresponding secondary IP addresses.
1204 Allows you to write a number, which can be used as required.
1207 xfrm4_gc_thresh - INTEGER
1208 The threshold at which we will start garbage collecting for IPv4
1209 destination cache entries. At twice this value the system will
1210 refuse new allocations.
1212 igmp_link_local_mcast_reports - BOOLEAN
1213 Enable IGMP reports for link local multicast groups in the
1218 kuznet@ms2.inr.ac.ru
1224 delon.nicolas@wanadoo.fr
1229 /proc/sys/net/ipv6/* Variables:
1231 IPv6 has no global variables such as tcp_*. tcp_* settings under ipv4/ also
1232 apply to IPv6 [XXX?].
1234 bindv6only - BOOLEAN
1235 Default value for IPV6_V6ONLY socket option,
1236 which restricts use of the IPv6 socket to IPv6 communication
1238 TRUE: disable IPv4-mapped address feature
1239 FALSE: enable IPv4-mapped address feature
1241 Default: FALSE (as specified in RFC3493)
1243 flowlabel_consistency - BOOLEAN
1244 Protect the consistency (and unicity) of flow label.
1245 You have to disable it to use IPV6_FL_F_REFLECT flag on the
1251 auto_flowlabels - INTEGER
1252 Automatically generate flow labels based on a flow hash of the
1253 packet. This allows intermediate devices, such as routers, to
1254 identify packet flows for mechanisms like Equal Cost Multipath
1255 Routing (see RFC 6438).
1256 0: automatic flow labels are completely disabled
1257 1: automatic flow labels are enabled by default, they can be
1258 disabled on a per socket basis using the IPV6_AUTOFLOWLABEL
1260 2: automatic flow labels are allowed, they may be enabled on a
1261 per socket basis using the IPV6_AUTOFLOWLABEL socket option
1262 3: automatic flow labels are enabled and enforced, they cannot
1263 be disabled by the socket option
1266 flowlabel_state_ranges - BOOLEAN
1267 Split the flow label number space into two ranges. 0-0x7FFFF is
1268 reserved for the IPv6 flow manager facility, 0x80000-0xFFFFF
1269 is reserved for stateless flow labels as described in RFC6437.
1274 anycast_src_echo_reply - BOOLEAN
1275 Controls the use of anycast addresses as source addresses for ICMPv6
1281 idgen_delay - INTEGER
1282 Controls the delay in seconds after which time to retry
1283 privacy stable address generation if a DAD conflict is
1285 Default: 1 (as specified in RFC7217)
1287 idgen_retries - INTEGER
1288 Controls the number of retries to generate a stable privacy
1289 address if a DAD conflict is detected.
1290 Default: 3 (as specified in RFC7217)
1293 Controls the MLD query robustness variable (see RFC3810 9.1).
1294 Default: 2 (as specified by RFC3810 9.1)
1295 Minimum: 1 (as specified by RFC6636 4.5)
1299 ip6frag_high_thresh - INTEGER
1300 Maximum memory used to reassemble IPv6 fragments. When
1301 ip6frag_high_thresh bytes of memory is allocated for this purpose,
1302 the fragment handler will toss packets until ip6frag_low_thresh
1305 ip6frag_low_thresh - INTEGER
1306 See ip6frag_high_thresh
1308 ip6frag_time - INTEGER
1309 Time in seconds to keep an IPv6 fragment in memory.
1312 Change the interface-specific default settings.
1316 Change all the interface-specific settings.
1318 [XXX: Other special features than forwarding?]
1320 conf/all/forwarding - BOOLEAN
1321 Enable global IPv6 forwarding between all interfaces.
1323 IPv4 and IPv6 work differently here; e.g. netfilter must be used
1324 to control which interfaces may forward packets and which not.
1326 This also sets all interfaces' Host/Router setting
1327 'forwarding' to the specified value. See below for details.
1329 This referred to as global forwarding.
1334 fwmark_reflect - BOOLEAN
1335 Controls the fwmark of kernel-generated IPv6 reply packets that are not
1336 associated with a socket for example, TCP RSTs or ICMPv6 echo replies).
1337 If unset, these packets have a fwmark of zero. If set, they have the
1338 fwmark of the packet they are replying to.
1342 Change special settings per interface.
1344 The functional behaviour for certain settings is different
1345 depending on whether local forwarding is enabled or not.
1348 Accept Router Advertisements; autoconfigure using them.
1350 It also determines whether or not to transmit Router
1351 Solicitations. If and only if the functional setting is to
1352 accept Router Advertisements, Router Solicitations will be
1355 Possible values are:
1356 0 Do not accept Router Advertisements.
1357 1 Accept Router Advertisements if forwarding is disabled.
1358 2 Overrule forwarding behaviour. Accept Router Advertisements
1359 even if forwarding is enabled.
1361 Functional default: enabled if local forwarding is disabled.
1362 disabled if local forwarding is enabled.
1364 accept_ra_defrtr - BOOLEAN
1365 Learn default router in Router Advertisement.
1367 Functional default: enabled if accept_ra is enabled.
1368 disabled if accept_ra is disabled.
1370 accept_ra_from_local - BOOLEAN
1371 Accept RA with source-address that is found on local machine
1372 if the RA is otherwise proper and able to be accepted.
1373 Default is to NOT accept these as it may be an un-intended
1377 enabled if accept_ra_from_local is enabled
1378 on a specific interface.
1379 disabled if accept_ra_from_local is disabled
1380 on a specific interface.
1382 accept_ra_min_hop_limit - INTEGER
1383 Minimum hop limit Information in Router Advertisement.
1385 Hop limit Information in Router Advertisement less than this
1386 variable shall be ignored.
1390 accept_ra_pinfo - BOOLEAN
1391 Learn Prefix Information in Router Advertisement.
1393 Functional default: enabled if accept_ra is enabled.
1394 disabled if accept_ra is disabled.
1396 accept_ra_rt_info_max_plen - INTEGER
1397 Maximum prefix length of Route Information in RA.
1399 Route Information w/ prefix larger than or equal to this
1400 variable shall be ignored.
1402 Functional default: 0 if accept_ra_rtr_pref is enabled.
1403 -1 if accept_ra_rtr_pref is disabled.
1405 accept_ra_rtr_pref - BOOLEAN
1406 Accept Router Preference in RA.
1408 Functional default: enabled if accept_ra is enabled.
1409 disabled if accept_ra is disabled.
1411 accept_ra_mtu - BOOLEAN
1412 Apply the MTU value specified in RA option 5 (RFC4861). If
1413 disabled, the MTU specified in the RA will be ignored.
1415 Functional default: enabled if accept_ra is enabled.
1416 disabled if accept_ra is disabled.
1418 accept_redirects - BOOLEAN
1421 Functional default: enabled if local forwarding is disabled.
1422 disabled if local forwarding is enabled.
1424 accept_source_route - INTEGER
1425 Accept source routing (routing extension header).
1427 >= 0: Accept only routing header type 2.
1428 < 0: Do not accept routing header.
1433 Autoconfigure addresses using Prefix Information in Router
1436 Functional default: enabled if accept_ra_pinfo is enabled.
1437 disabled if accept_ra_pinfo is disabled.
1439 dad_transmits - INTEGER
1440 The amount of Duplicate Address Detection probes to send.
1443 forwarding - INTEGER
1444 Configure interface-specific Host/Router behaviour.
1446 Note: It is recommended to have the same setting on all
1447 interfaces; mixed router/host scenarios are rather uncommon.
1449 Possible values are:
1450 0 Forwarding disabled
1451 1 Forwarding enabled
1455 By default, Host behaviour is assumed. This means:
1457 1. IsRouter flag is not set in Neighbour Advertisements.
1458 2. If accept_ra is TRUE (default), transmit Router
1460 3. If accept_ra is TRUE (default), accept Router
1461 Advertisements (and do autoconfiguration).
1462 4. If accept_redirects is TRUE (default), accept Redirects.
1466 If local forwarding is enabled, Router behaviour is assumed.
1467 This means exactly the reverse from the above:
1469 1. IsRouter flag is set in Neighbour Advertisements.
1470 2. Router Solicitations are not sent unless accept_ra is 2.
1471 3. Router Advertisements are ignored unless accept_ra is 2.
1472 4. Redirects are ignored.
1474 Default: 0 (disabled) if global forwarding is disabled (default),
1475 otherwise 1 (enabled).
1478 Default Hop Limit to set.
1482 Default Maximum Transfer Unit
1483 Default: 1280 (IPv6 required minimum)
1485 ip_nonlocal_bind - BOOLEAN
1486 If set, allows processes to bind() to non-local IPv6 addresses,
1487 which can be quite useful - but may break some applications.
1490 router_probe_interval - INTEGER
1491 Minimum interval (in seconds) between Router Probing described
1496 router_solicitation_delay - INTEGER
1497 Number of seconds to wait after interface is brought up
1498 before sending Router Solicitations.
1501 router_solicitation_interval - INTEGER
1502 Number of seconds to wait between Router Solicitations.
1505 router_solicitations - INTEGER
1506 Number of Router Solicitations to send until assuming no
1507 routers are present.
1510 use_oif_addrs_only - BOOLEAN
1511 When enabled, the candidate source addresses for destinations
1512 routed via this interface are restricted to the set of addresses
1513 configured on this interface (vis. RFC 6724, section 4).
1517 use_tempaddr - INTEGER
1518 Preference for Privacy Extensions (RFC3041).
1519 <= 0 : disable Privacy Extensions
1520 == 1 : enable Privacy Extensions, but prefer public
1521 addresses over temporary addresses.
1522 > 1 : enable Privacy Extensions and prefer temporary
1523 addresses over public addresses.
1524 Default: 0 (for most devices)
1525 -1 (for point-to-point devices and loopback devices)
1527 temp_valid_lft - INTEGER
1528 valid lifetime (in seconds) for temporary addresses.
1529 Default: 604800 (7 days)
1531 temp_prefered_lft - INTEGER
1532 Preferred lifetime (in seconds) for temporary addresses.
1533 Default: 86400 (1 day)
1535 max_desync_factor - INTEGER
1536 Maximum value for DESYNC_FACTOR, which is a random value
1537 that ensures that clients don't synchronize with each
1538 other and generate new addresses at exactly the same time.
1539 value is in seconds.
1542 regen_max_retry - INTEGER
1543 Number of attempts before give up attempting to generate
1544 valid temporary addresses.
1547 max_addresses - INTEGER
1548 Maximum number of autoconfigured addresses per interface. Setting
1549 to zero disables the limitation. It is not recommended to set this
1550 value too large (or to zero) because it would be an easy way to
1551 crash the kernel by allowing too many addresses to be created.
1554 disable_ipv6 - BOOLEAN
1555 Disable IPv6 operation. If accept_dad is set to 2, this value
1556 will be dynamically set to TRUE if DAD fails for the link-local
1558 Default: FALSE (enable IPv6 operation)
1560 When this value is changed from 1 to 0 (IPv6 is being enabled),
1561 it will dynamically create a link-local address on the given
1562 interface and start Duplicate Address Detection, if necessary.
1564 When this value is changed from 0 to 1 (IPv6 is being disabled),
1565 it will dynamically delete all address on the given interface.
1567 accept_dad - INTEGER
1568 Whether to accept DAD (Duplicate Address Detection).
1570 1: Enable DAD (default)
1571 2: Enable DAD, and disable IPv6 operation if MAC-based duplicate
1572 link-local address has been found.
1574 force_tllao - BOOLEAN
1575 Enable sending the target link-layer address option even when
1576 responding to a unicast neighbor solicitation.
1579 Quoting from RFC 2461, section 4.4, Target link-layer address:
1581 "The option MUST be included for multicast solicitations in order to
1582 avoid infinite Neighbor Solicitation "recursion" when the peer node
1583 does not have a cache entry to return a Neighbor Advertisements
1584 message. When responding to unicast solicitations, the option can be
1585 omitted since the sender of the solicitation has the correct link-
1586 layer address; otherwise it would not have be able to send the unicast
1587 solicitation in the first place. However, including the link-layer
1588 address in this case adds little overhead and eliminates a potential
1589 race condition where the sender deletes the cached link-layer address
1590 prior to receiving a response to a previous solicitation."
1592 ndisc_notify - BOOLEAN
1593 Define mode for notification of address and device changes.
1594 0 - (default): do nothing
1595 1 - Generate unsolicited neighbour advertisements when device is brought
1596 up or hardware address changes.
1598 mldv1_unsolicited_report_interval - INTEGER
1599 The interval in milliseconds in which the next unsolicited
1600 MLDv1 report retransmit will take place.
1601 Default: 10000 (10 seconds)
1603 mldv2_unsolicited_report_interval - INTEGER
1604 The interval in milliseconds in which the next unsolicited
1605 MLDv2 report retransmit will take place.
1606 Default: 1000 (1 second)
1608 force_mld_version - INTEGER
1609 0 - (default) No enforcement of a MLD version, MLDv1 fallback allowed
1610 1 - Enforce to use MLD version 1
1611 2 - Enforce to use MLD version 2
1613 suppress_frag_ndisc - INTEGER
1614 Control RFC 6980 (Security Implications of IPv6 Fragmentation
1615 with IPv6 Neighbor Discovery) behavior:
1616 1 - (default) discard fragmented neighbor discovery packets
1617 0 - allow fragmented neighbor discovery packets
1619 optimistic_dad - BOOLEAN
1620 Whether to perform Optimistic Duplicate Address Detection (RFC 4429).
1621 0: disabled (default)
1624 use_optimistic - BOOLEAN
1625 If enabled, do not classify optimistic addresses as deprecated during
1626 source address selection. Preferred addresses will still be chosen
1627 before optimistic addresses, subject to other ranking in the source
1628 address selection algorithm.
1629 0: disabled (default)
1632 stable_secret - IPv6 address
1633 This IPv6 address will be used as a secret to generate IPv6
1634 addresses for link-local addresses and autoconfigured
1635 ones. All addresses generated after setting this secret will
1636 be stable privacy ones by default. This can be changed via the
1637 addrgenmode ip-link. conf/default/stable_secret is used as the
1638 secret for the namespace, the interface specific ones can
1639 overwrite that. Writes to conf/all/stable_secret are refused.
1641 It is recommended to generate this secret during installation
1642 of a system and keep it stable after that.
1644 By default the stable secret is unset.
1648 Limit the maximal rates for sending ICMPv6 packets.
1649 0 to disable any limiting,
1650 otherwise the minimal space between responses in milliseconds.
1653 xfrm6_gc_thresh - INTEGER
1654 The threshold at which we will start garbage collecting for IPv6
1655 destination cache entries. At twice this value the system will
1656 refuse new allocations.
1660 Pekka Savola <pekkas@netcore.fi>
1661 YOSHIFUJI Hideaki / USAGI Project <yoshfuji@linux-ipv6.org>
1664 /proc/sys/net/bridge/* Variables:
1666 bridge-nf-call-arptables - BOOLEAN
1667 1 : pass bridged ARP traffic to arptables' FORWARD chain.
1671 bridge-nf-call-iptables - BOOLEAN
1672 1 : pass bridged IPv4 traffic to iptables' chains.
1676 bridge-nf-call-ip6tables - BOOLEAN
1677 1 : pass bridged IPv6 traffic to ip6tables' chains.
1681 bridge-nf-filter-vlan-tagged - BOOLEAN
1682 1 : pass bridged vlan-tagged ARP/IP/IPv6 traffic to {arp,ip,ip6}tables.
1686 bridge-nf-filter-pppoe-tagged - BOOLEAN
1687 1 : pass bridged pppoe-tagged IP/IPv6 traffic to {ip,ip6}tables.
1691 bridge-nf-pass-vlan-input-dev - BOOLEAN
1692 1: if bridge-nf-filter-vlan-tagged is enabled, try to find a vlan
1693 interface on the bridge and set the netfilter input device to the vlan.
1694 This allows use of e.g. "iptables -i br0.1" and makes the REDIRECT
1695 target work with vlan-on-top-of-bridge interfaces. When no matching
1696 vlan interface is found, or this switch is off, the input device is
1697 set to the bridge interface.
1698 0: disable bridge netfilter vlan interface lookup.
1701 proc/sys/net/sctp/* Variables:
1703 addip_enable - BOOLEAN
1704 Enable or disable extension of Dynamic Address Reconfiguration
1705 (ADD-IP) functionality specified in RFC5061. This extension provides
1706 the ability to dynamically add and remove new addresses for the SCTP
1709 1: Enable extension.
1711 0: Disable extension.
1715 addip_noauth_enable - BOOLEAN
1716 Dynamic Address Reconfiguration (ADD-IP) requires the use of
1717 authentication to protect the operations of adding or removing new
1718 addresses. This requirement is mandated so that unauthorized hosts
1719 would not be able to hijack associations. However, older
1720 implementations may not have implemented this requirement while
1721 allowing the ADD-IP extension. For reasons of interoperability,
1722 we provide this variable to control the enforcement of the
1723 authentication requirement.
1725 1: Allow ADD-IP extension to be used without authentication. This
1726 should only be set in a closed environment for interoperability
1727 with older implementations.
1729 0: Enforce the authentication requirement
1733 auth_enable - BOOLEAN
1734 Enable or disable Authenticated Chunks extension. This extension
1735 provides the ability to send and receive authenticated chunks and is
1736 required for secure operation of Dynamic Address Reconfiguration
1739 1: Enable this extension.
1740 0: Disable this extension.
1744 prsctp_enable - BOOLEAN
1745 Enable or disable the Partial Reliability extension (RFC3758) which
1746 is used to notify peers that a given DATA should no longer be expected.
1754 The limit of the number of new packets that can be initially sent. It
1755 controls how bursty the generated traffic can be.
1759 association_max_retrans - INTEGER
1760 Set the maximum number for retransmissions that an association can
1761 attempt deciding that the remote end is unreachable. If this value
1762 is exceeded, the association is terminated.
1766 max_init_retransmits - INTEGER
1767 The maximum number of retransmissions of INIT and COOKIE-ECHO chunks
1768 that an association will attempt before declaring the destination
1769 unreachable and terminating.
1773 path_max_retrans - INTEGER
1774 The maximum number of retransmissions that will be attempted on a given
1775 path. Once this threshold is exceeded, the path is considered
1776 unreachable, and new traffic will use a different path when the
1777 association is multihomed.
1781 pf_retrans - INTEGER
1782 The number of retransmissions that will be attempted on a given path
1783 before traffic is redirected to an alternate transport (should one
1784 exist). Note this is distinct from path_max_retrans, as a path that
1785 passes the pf_retrans threshold can still be used. Its only
1786 deprioritized when a transmission path is selected by the stack. This
1787 setting is primarily used to enable fast failover mechanisms without
1788 having to reduce path_max_retrans to a very low value. See:
1789 http://www.ietf.org/id/draft-nishida-tsvwg-sctp-failover-05.txt
1790 for details. Note also that a value of pf_retrans > path_max_retrans
1791 disables this feature
1795 rto_initial - INTEGER
1796 The initial round trip timeout value in milliseconds that will be used
1797 in calculating round trip times. This is the initial time interval
1798 for retransmissions.
1803 The maximum value (in milliseconds) of the round trip timeout. This
1804 is the largest time interval that can elapse between retransmissions.
1809 The minimum value (in milliseconds) of the round trip timeout. This
1810 is the smallest time interval the can elapse between retransmissions.
1814 hb_interval - INTEGER
1815 The interval (in milliseconds) between HEARTBEAT chunks. These chunks
1816 are sent at the specified interval on idle paths to probe the state of
1817 a given path between 2 associations.
1821 sack_timeout - INTEGER
1822 The amount of time (in milliseconds) that the implementation will wait
1827 valid_cookie_life - INTEGER
1828 The default lifetime of the SCTP cookie (in milliseconds). The cookie
1829 is used during association establishment.
1833 cookie_preserve_enable - BOOLEAN
1834 Enable or disable the ability to extend the lifetime of the SCTP cookie
1835 that is used during the establishment phase of SCTP association
1837 1: Enable cookie lifetime extension.
1842 cookie_hmac_alg - STRING
1843 Select the hmac algorithm used when generating the cookie value sent by
1844 a listening sctp socket to a connecting client in the INIT-ACK chunk.
1849 Ability to assign md5 or sha1 as the selected alg is predicated on the
1850 configuration of those algorithms at build time (CONFIG_CRYPTO_MD5 and
1851 CONFIG_CRYPTO_SHA1).
1853 Default: Dependent on configuration. MD5 if available, else SHA1 if
1854 available, else none.
1856 rcvbuf_policy - INTEGER
1857 Determines if the receive buffer is attributed to the socket or to
1858 association. SCTP supports the capability to create multiple
1859 associations on a single socket. When using this capability, it is
1860 possible that a single stalled association that's buffering a lot
1861 of data may block other associations from delivering their data by
1862 consuming all of the receive buffer space. To work around this,
1863 the rcvbuf_policy could be set to attribute the receiver buffer space
1864 to each association instead of the socket. This prevents the described
1867 1: rcvbuf space is per association
1868 0: rcvbuf space is per socket
1872 sndbuf_policy - INTEGER
1873 Similar to rcvbuf_policy above, this applies to send buffer space.
1875 1: Send buffer is tracked per association
1876 0: Send buffer is tracked per socket.
1880 sctp_mem - vector of 3 INTEGERs: min, pressure, max
1881 Number of pages allowed for queueing by all SCTP sockets.
1883 min: Below this number of pages SCTP is not bothered about its
1884 memory appetite. When amount of memory allocated by SCTP exceeds
1885 this number, SCTP starts to moderate memory usage.
1887 pressure: This value was introduced to follow format of tcp_mem.
1889 max: Number of pages allowed for queueing by all SCTP sockets.
1891 Default is calculated at boot time from amount of available memory.
1893 sctp_rmem - vector of 3 INTEGERs: min, default, max
1894 Only the first value ("min") is used, "default" and "max" are
1897 min: Minimal size of receive buffer used by SCTP socket.
1898 It is guaranteed to each SCTP socket (but not association) even
1899 under moderate memory pressure.
1903 sctp_wmem - vector of 3 INTEGERs: min, default, max
1904 Currently this tunable has no effect.
1906 addr_scope_policy - INTEGER
1907 Control IPv4 address scoping - draft-stewart-tsvwg-sctp-ipv4-00
1909 0 - Disable IPv4 address scoping
1910 1 - Enable IPv4 address scoping
1911 2 - Follow draft but allow IPv4 private addresses
1912 3 - Follow draft but allow IPv4 link local addresses
1917 /proc/sys/net/core/*
1918 Please see: Documentation/sysctl/net.txt for descriptions of these entries.
1921 /proc/sys/net/unix/*
1922 max_dgram_qlen - INTEGER
1923 The maximum length of dgram socket receive queue
1930 /proc/sys/net/irda/*
1931 fast_poll_increase FIXME
1932 warn_noreply_time FIXME
1933 discovery_slots FIXME
1936 discovery_timeout FIXME
1937 lap_keepalive_time FIXME
1938 max_noreply_time FIXME
1939 max_tx_data_size FIXME
1941 min_tx_turn_time FIXME