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 hardened 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 fib_multipath_use_neigh - BOOLEAN
67 Use status of existing neighbor entry when determining nexthop for
68 multipath routes. If disabled, neighbor information is not used and
69 packets could be directed to a failed nexthop. Only valid for kernels
70 built with CONFIG_IP_ROUTE_MULTIPATH enabled.
76 fib_multipath_hash_policy - INTEGER
77 Controls which hash policy to use for multipath routes. Only valid
78 for kernels built with CONFIG_IP_ROUTE_MULTIPATH enabled.
84 ip_forward_update_priority - INTEGER
85 Whether to update SKB priority from "TOS" field in IPv4 header after it
86 is forwarded. The new SKB priority is mapped from TOS field value
87 according to an rt_tos2priority table (see e.g. man tc-prio).
88 Default: 1 (Update priority.)
90 0 - Do not update priority.
93 route/max_size - INTEGER
94 Maximum number of routes allowed in the kernel. Increase
95 this when using large numbers of interfaces and/or routes.
96 From linux kernel 3.6 onwards, this is deprecated for ipv4
97 as route cache is no longer used.
99 neigh/default/gc_thresh1 - INTEGER
100 Minimum number of entries to keep. Garbage collector will not
101 purge entries if there are fewer than this number.
104 neigh/default/gc_thresh2 - INTEGER
105 Threshold when garbage collector becomes more aggressive about
106 purging entries. Entries older than 5 seconds will be cleared
107 when over this number.
110 neigh/default/gc_thresh3 - INTEGER
111 Maximum number of non-PERMANENT neighbor entries allowed. Increase
112 this when using large numbers of interfaces and when communicating
113 with large numbers of directly-connected peers.
116 neigh/default/unres_qlen_bytes - INTEGER
117 The maximum number of bytes which may be used by packets
118 queued for each unresolved address by other network layers.
120 Setting negative value is meaningless and will return error.
121 Default: SK_WMEM_MAX, (same as net.core.wmem_default).
122 Exact value depends on architecture and kernel options,
123 but should be enough to allow queuing 256 packets
126 neigh/default/unres_qlen - INTEGER
127 The maximum number of packets which may be queued for each
128 unresolved address by other network layers.
129 (deprecated in linux 3.3) : use unres_qlen_bytes instead.
130 Prior to linux 3.3, the default value is 3 which may cause
131 unexpected packet loss. The current default value is calculated
132 according to default value of unres_qlen_bytes and true size of
136 mtu_expires - INTEGER
137 Time, in seconds, that cached PMTU information is kept.
139 min_adv_mss - INTEGER
140 The advertised MSS depends on the first hop route MTU, but will
141 never be lower than this setting.
145 ipfrag_high_thresh - LONG INTEGER
146 Maximum memory used to reassemble IP fragments.
148 ipfrag_low_thresh - LONG INTEGER
149 (Obsolete since linux-4.17)
150 Maximum memory used to reassemble IP fragments before the kernel
151 begins to remove incomplete fragment queues to free up resources.
152 The kernel still accepts new fragments for defragmentation.
154 ipfrag_time - INTEGER
155 Time in seconds to keep an IP fragment in memory.
157 ipfrag_max_dist - INTEGER
158 ipfrag_max_dist is a non-negative integer value which defines the
159 maximum "disorder" which is allowed among fragments which share a
160 common IP source address. Note that reordering of packets is
161 not unusual, but if a large number of fragments arrive from a source
162 IP address while a particular fragment queue remains incomplete, it
163 probably indicates that one or more fragments belonging to that queue
164 have been lost. When ipfrag_max_dist is positive, an additional check
165 is done on fragments before they are added to a reassembly queue - if
166 ipfrag_max_dist (or more) fragments have arrived from a particular IP
167 address between additions to any IP fragment queue using that source
168 address, it's presumed that one or more fragments in the queue are
169 lost. The existing fragment queue will be dropped, and a new one
170 started. An ipfrag_max_dist value of zero disables this check.
172 Using a very small value, e.g. 1 or 2, for ipfrag_max_dist can
173 result in unnecessarily dropping fragment queues when normal
174 reordering of packets occurs, which could lead to poor application
175 performance. Using a very large value, e.g. 50000, increases the
176 likelihood of incorrectly reassembling IP fragments that originate
177 from different IP datagrams, which could result in data corruption.
182 inet_peer_threshold - INTEGER
183 The approximate size of the storage. Starting from this threshold
184 entries will be thrown aggressively. This threshold also determines
185 entries' time-to-live and time intervals between garbage collection
186 passes. More entries, less time-to-live, less GC interval.
188 inet_peer_minttl - INTEGER
189 Minimum time-to-live of entries. Should be enough to cover fragment
190 time-to-live on the reassembling side. This minimum time-to-live is
191 guaranteed if the pool size is less than inet_peer_threshold.
194 inet_peer_maxttl - INTEGER
195 Maximum time-to-live of entries. Unused entries will expire after
196 this period of time if there is no memory pressure on the pool (i.e.
197 when the number of entries in the pool is very small).
203 Limit of socket listen() backlog, known in userspace as SOMAXCONN.
204 Defaults to 128. See also tcp_max_syn_backlog for additional tuning
207 tcp_abort_on_overflow - BOOLEAN
208 If listening service is too slow to accept new connections,
209 reset them. Default state is FALSE. It means that if overflow
210 occurred due to a burst, connection will recover. Enable this
211 option _only_ if you are really sure that listening daemon
212 cannot be tuned to accept connections faster. Enabling this
213 option can harm clients of your server.
215 tcp_adv_win_scale - INTEGER
216 Count buffering overhead as bytes/2^tcp_adv_win_scale
217 (if tcp_adv_win_scale > 0) or bytes-bytes/2^(-tcp_adv_win_scale),
219 Possible values are [-31, 31], inclusive.
222 tcp_allowed_congestion_control - STRING
223 Show/set the congestion control choices available to non-privileged
224 processes. The list is a subset of those listed in
225 tcp_available_congestion_control.
226 Default is "reno" and the default setting (tcp_congestion_control).
228 tcp_app_win - INTEGER
229 Reserve max(window/2^tcp_app_win, mss) of window for application
230 buffer. Value 0 is special, it means that nothing is reserved.
233 tcp_autocorking - BOOLEAN
234 Enable TCP auto corking :
235 When applications do consecutive small write()/sendmsg() system calls,
236 we try to coalesce these small writes as much as possible, to lower
237 total amount of sent packets. This is done if at least one prior
238 packet for the flow is waiting in Qdisc queues or device transmit
239 queue. Applications can still use TCP_CORK for optimal behavior
240 when they know how/when to uncork their sockets.
243 tcp_available_congestion_control - STRING
244 Shows the available congestion control choices that are registered.
245 More congestion control algorithms may be available as modules,
248 tcp_base_mss - INTEGER
249 The initial value of search_low to be used by the packetization layer
250 Path MTU discovery (MTU probing). If MTU probing is enabled,
251 this is the initial MSS used by the connection.
253 tcp_congestion_control - STRING
254 Set the congestion control algorithm to be used for new
255 connections. The algorithm "reno" is always available, but
256 additional choices may be available based on kernel configuration.
257 Default is set as part of kernel configuration.
258 For passive connections, the listener congestion control choice
260 [see setsockopt(listenfd, SOL_TCP, TCP_CONGESTION, "name" ...) ]
263 Allows TCP to send "duplicate" SACKs.
265 tcp_early_retrans - INTEGER
266 Tail loss probe (TLP) converts RTOs occurring due to tail
267 losses into fast recovery (draft-ietf-tcpm-rack). Note that
268 TLP requires RACK to function properly (see tcp_recovery below)
275 Control use of Explicit Congestion Notification (ECN) by TCP.
276 ECN is used only when both ends of the TCP connection indicate
277 support for it. This feature is useful in avoiding losses due
278 to congestion by allowing supporting routers to signal
279 congestion before having to drop packets.
281 0 Disable ECN. Neither initiate nor accept ECN.
282 1 Enable ECN when requested by incoming connections and
283 also request ECN on outgoing connection attempts.
284 2 Enable ECN when requested by incoming connections
285 but do not request ECN on outgoing connections.
288 tcp_ecn_fallback - BOOLEAN
289 If the kernel detects that ECN connection misbehaves, enable fall
290 back to non-ECN. Currently, this knob implements the fallback
291 from RFC3168, section 6.1.1.1., but we reserve that in future,
292 additional detection mechanisms could be implemented under this
293 knob. The value is not used, if tcp_ecn or per route (or congestion
294 control) ECN settings are disabled.
295 Default: 1 (fallback enabled)
298 This is a legacy option, it has no effect anymore.
300 tcp_fin_timeout - INTEGER
301 The length of time an orphaned (no longer referenced by any
302 application) connection will remain in the FIN_WAIT_2 state
303 before it is aborted at the local end. While a perfectly
304 valid "receive only" state for an un-orphaned connection, an
305 orphaned connection in FIN_WAIT_2 state could otherwise wait
306 forever for the remote to close its end of the connection.
311 Enables Forward RTO-Recovery (F-RTO) defined in RFC5682.
312 F-RTO is an enhanced recovery algorithm for TCP retransmission
313 timeouts. It is particularly beneficial in networks where the
314 RTT fluctuates (e.g., wireless). F-RTO is sender-side only
315 modification. It does not require any support from the peer.
317 By default it's enabled with a non-zero value. 0 disables F-RTO.
319 tcp_fwmark_accept - BOOLEAN
320 If set, incoming connections to listening sockets that do not have a
321 socket mark will set the mark of the accepting socket to the fwmark of
322 the incoming SYN packet. This will cause all packets on that connection
323 (starting from the first SYNACK) to be sent with that fwmark. The
324 listening socket's mark is unchanged. Listening sockets that already
325 have a fwmark set via setsockopt(SOL_SOCKET, SO_MARK, ...) are
330 tcp_invalid_ratelimit - INTEGER
331 Limit the maximal rate for sending duplicate acknowledgments
332 in response to incoming TCP packets that are for an existing
333 connection but that are invalid due to any of these reasons:
335 (a) out-of-window sequence number,
336 (b) out-of-window acknowledgment number, or
337 (c) PAWS (Protection Against Wrapped Sequence numbers) check failure
339 This can help mitigate simple "ack loop" DoS attacks, wherein
340 a buggy or malicious middlebox or man-in-the-middle can
341 rewrite TCP header fields in manner that causes each endpoint
342 to think that the other is sending invalid TCP segments, thus
343 causing each side to send an unterminating stream of duplicate
344 acknowledgments for invalid segments.
346 Using 0 disables rate-limiting of dupacks in response to
347 invalid segments; otherwise this value specifies the minimal
348 space between sending such dupacks, in milliseconds.
350 Default: 500 (milliseconds).
352 tcp_keepalive_time - INTEGER
353 How often TCP sends out keepalive messages when keepalive is enabled.
356 tcp_keepalive_probes - INTEGER
357 How many keepalive probes TCP sends out, until it decides that the
358 connection is broken. Default value: 9.
360 tcp_keepalive_intvl - INTEGER
361 How frequently the probes are send out. Multiplied by
362 tcp_keepalive_probes it is time to kill not responding connection,
363 after probes started. Default value: 75sec i.e. connection
364 will be aborted after ~11 minutes of retries.
366 tcp_l3mdev_accept - BOOLEAN
367 Enables child sockets to inherit the L3 master device index.
368 Enabling this option allows a "global" listen socket to work
369 across L3 master domains (e.g., VRFs) with connected sockets
370 derived from the listen socket to be bound to the L3 domain in
371 which the packets originated. Only valid when the kernel was
372 compiled with CONFIG_NET_L3_MASTER_DEV.
373 Default: 0 (disabled)
375 tcp_low_latency - BOOLEAN
376 This is a legacy option, it has no effect anymore.
378 tcp_max_orphans - INTEGER
379 Maximal number of TCP sockets not attached to any user file handle,
380 held by system. If this number is exceeded orphaned connections are
381 reset immediately and warning is printed. This limit exists
382 only to prevent simple DoS attacks, you _must_ not rely on this
383 or lower the limit artificially, but rather increase it
384 (probably, after increasing installed memory),
385 if network conditions require more than default value,
386 and tune network services to linger and kill such states
387 more aggressively. Let me to remind again: each orphan eats
388 up to ~64K of unswappable memory.
390 tcp_max_syn_backlog - INTEGER
391 Maximal number of remembered connection requests, which have not
392 received an acknowledgment from connecting client.
393 The minimal value is 128 for low memory machines, and it will
394 increase in proportion to the memory of machine.
395 If server suffers from overload, try increasing this number.
397 tcp_max_tw_buckets - INTEGER
398 Maximal number of timewait sockets held by system simultaneously.
399 If this number is exceeded time-wait socket is immediately destroyed
400 and warning is printed. This limit exists only to prevent
401 simple DoS attacks, you _must_ not lower the limit artificially,
402 but rather increase it (probably, after increasing installed memory),
403 if network conditions require more than default value.
405 tcp_mem - vector of 3 INTEGERs: min, pressure, max
406 min: below this number of pages TCP is not bothered about its
409 pressure: when amount of memory allocated by TCP exceeds this number
410 of pages, TCP moderates its memory consumption and enters memory
411 pressure mode, which is exited when memory consumption falls
414 max: number of pages allowed for queueing by all TCP sockets.
416 Defaults are calculated at boot time from amount of available
419 tcp_min_rtt_wlen - INTEGER
420 The window length of the windowed min filter to track the minimum RTT.
421 A shorter window lets a flow more quickly pick up new (higher)
422 minimum RTT when it is moved to a longer path (e.g., due to traffic
423 engineering). A longer window makes the filter more resistant to RTT
424 inflations such as transient congestion. The unit is seconds.
427 tcp_moderate_rcvbuf - BOOLEAN
428 If set, TCP performs receive buffer auto-tuning, attempting to
429 automatically size the buffer (no greater than tcp_rmem[2]) to
430 match the size required by the path for full throughput. Enabled by
433 tcp_mtu_probing - INTEGER
434 Controls TCP Packetization-Layer Path MTU Discovery. Takes three
437 1 - Disabled by default, enabled when an ICMP black hole detected
438 2 - Always enabled, use initial MSS of tcp_base_mss.
440 tcp_probe_interval - UNSIGNED INTEGER
441 Controls how often to start TCP Packetization-Layer Path MTU
442 Discovery reprobe. The default is reprobing every 10 minutes as
445 tcp_probe_threshold - INTEGER
446 Controls when TCP Packetization-Layer Path MTU Discovery probing
447 will stop in respect to the width of search range in bytes. Default
450 tcp_no_metrics_save - BOOLEAN
451 By default, TCP saves various connection metrics in the route cache
452 when the connection closes, so that connections established in the
453 near future can use these to set initial conditions. Usually, this
454 increases overall performance, but may sometimes cause performance
455 degradation. If set, TCP will not cache metrics on closing
458 tcp_orphan_retries - INTEGER
459 This value influences the timeout of a locally closed TCP connection,
460 when RTO retransmissions remain unacknowledged.
461 See tcp_retries2 for more details.
463 The default value is 8.
464 If your machine is a loaded WEB server,
465 you should think about lowering this value, such sockets
466 may consume significant resources. Cf. tcp_max_orphans.
468 tcp_recovery - INTEGER
469 This value is a bitmap to enable various experimental loss recovery
472 RACK: 0x1 enables the RACK loss detection for fast detection of lost
473 retransmissions and tail drops. It also subsumes and disables
474 RFC6675 recovery for SACK connections.
475 RACK: 0x2 makes RACK's reordering window static (min_rtt/4).
476 RACK: 0x4 disables RACK's DUPACK threshold heuristic
480 tcp_reordering - INTEGER
481 Initial reordering level of packets in a TCP stream.
482 TCP stack can then dynamically adjust flow reordering level
483 between this initial value and tcp_max_reordering
486 tcp_max_reordering - INTEGER
487 Maximal reordering level of packets in a TCP stream.
488 300 is a fairly conservative value, but you might increase it
489 if paths are using per packet load balancing (like bonding rr mode)
492 tcp_retrans_collapse - BOOLEAN
493 Bug-to-bug compatibility with some broken printers.
494 On retransmit try to send bigger packets to work around bugs in
497 tcp_retries1 - INTEGER
498 This value influences the time, after which TCP decides, that
499 something is wrong due to unacknowledged RTO retransmissions,
500 and reports this suspicion to the network layer.
501 See tcp_retries2 for more details.
503 RFC 1122 recommends at least 3 retransmissions, which is the
506 tcp_retries2 - INTEGER
507 This value influences the timeout of an alive TCP connection,
508 when RTO retransmissions remain unacknowledged.
509 Given a value of N, a hypothetical TCP connection following
510 exponential backoff with an initial RTO of TCP_RTO_MIN would
511 retransmit N times before killing the connection at the (N+1)th RTO.
513 The default value of 15 yields a hypothetical timeout of 924.6
514 seconds and is a lower bound for the effective timeout.
515 TCP will effectively time out at the first RTO which exceeds the
516 hypothetical timeout.
518 RFC 1122 recommends at least 100 seconds for the timeout,
519 which corresponds to a value of at least 8.
521 tcp_rfc1337 - BOOLEAN
522 If set, the TCP stack behaves conforming to RFC1337. If unset,
523 we are not conforming to RFC, but prevent TCP TIME_WAIT
527 tcp_rmem - vector of 3 INTEGERs: min, default, max
528 min: Minimal size of receive buffer used by TCP sockets.
529 It is guaranteed to each TCP socket, even under moderate memory
533 default: initial size of receive buffer used by TCP sockets.
534 This value overrides net.core.rmem_default used by other protocols.
535 Default: 87380 bytes. This value results in window of 65535 with
536 default setting of tcp_adv_win_scale and tcp_app_win:0 and a bit
537 less for default tcp_app_win. See below about these variables.
539 max: maximal size of receive buffer allowed for automatically
540 selected receiver buffers for TCP socket. This value does not override
541 net.core.rmem_max. Calling setsockopt() with SO_RCVBUF disables
542 automatic tuning of that socket's receive buffer size, in which
543 case this value is ignored.
544 Default: between 87380B and 6MB, depending on RAM size.
547 Enable select acknowledgments (SACKS).
549 tcp_comp_sack_delay_ns - LONG INTEGER
550 TCP tries to reduce number of SACK sent, using a timer
551 based on 5% of SRTT, capped by this sysctl, in nano seconds.
552 The default is 1ms, based on TSO autosizing period.
554 Default : 1,000,000 ns (1 ms)
556 tcp_comp_sack_nr - INTEGER
557 Max numer of SACK that can be compressed.
558 Using 0 disables SACK compression.
562 tcp_slow_start_after_idle - BOOLEAN
563 If set, provide RFC2861 behavior and time out the congestion
564 window after an idle period. An idle period is defined at
565 the current RTO. If unset, the congestion window will not
566 be timed out after an idle period.
570 Use the Host requirements interpretation of the TCP urgent pointer field.
571 Most hosts use the older BSD interpretation, so if you turn this on
572 Linux might not communicate correctly with them.
575 tcp_synack_retries - INTEGER
576 Number of times SYNACKs for a passive TCP connection attempt will
577 be retransmitted. Should not be higher than 255. Default value
578 is 5, which corresponds to 31seconds till the last retransmission
579 with the current initial RTO of 1second. With this the final timeout
580 for a passive TCP connection will happen after 63seconds.
582 tcp_syncookies - BOOLEAN
583 Only valid when the kernel was compiled with CONFIG_SYN_COOKIES
584 Send out syncookies when the syn backlog queue of a socket
585 overflows. This is to prevent against the common 'SYN flood attack'
588 Note, that syncookies is fallback facility.
589 It MUST NOT be used to help highly loaded servers to stand
590 against legal connection rate. If you see SYN flood warnings
591 in your logs, but investigation shows that they occur
592 because of overload with legal connections, you should tune
593 another parameters until this warning disappear.
594 See: tcp_max_syn_backlog, tcp_synack_retries, tcp_abort_on_overflow.
596 syncookies seriously violate TCP protocol, do not allow
597 to use TCP extensions, can result in serious degradation
598 of some services (f.e. SMTP relaying), visible not by you,
599 but your clients and relays, contacting you. While you see
600 SYN flood warnings in logs not being really flooded, your server
601 is seriously misconfigured.
603 If you want to test which effects syncookies have to your
604 network connections you can set this knob to 2 to enable
605 unconditionally generation of syncookies.
607 tcp_fastopen - INTEGER
608 Enable TCP Fast Open (RFC7413) to send and accept data in the opening
611 The client support is enabled by flag 0x1 (on by default). The client
612 then must use sendmsg() or sendto() with the MSG_FASTOPEN flag,
613 rather than connect() to send data in SYN.
615 The server support is enabled by flag 0x2 (off by default). Then
616 either enable for all listeners with another flag (0x400) or
617 enable individual listeners via TCP_FASTOPEN socket option with
618 the option value being the length of the syn-data backlog.
620 The values (bitmap) are
621 0x1: (client) enables sending data in the opening SYN on the client.
622 0x2: (server) enables the server support, i.e., allowing data in
623 a SYN packet to be accepted and passed to the
624 application before 3-way handshake finishes.
625 0x4: (client) send data in the opening SYN regardless of cookie
626 availability and without a cookie option.
627 0x200: (server) accept data-in-SYN w/o any cookie option present.
628 0x400: (server) enable all listeners to support Fast Open by
629 default without explicit TCP_FASTOPEN socket option.
633 Note that that additional client or server features are only
634 effective if the basic support (0x1 and 0x2) are enabled respectively.
636 tcp_fastopen_blackhole_timeout_sec - INTEGER
637 Initial time period in second to disable Fastopen on active TCP sockets
638 when a TFO firewall blackhole issue happens.
639 This time period will grow exponentially when more blackhole issues
640 get detected right after Fastopen is re-enabled and will reset to
641 initial value when the blackhole issue goes away.
642 0 to disable the blackhole detection.
643 By default, it is set to 1hr.
645 tcp_syn_retries - INTEGER
646 Number of times initial SYNs for an active TCP connection attempt
647 will be retransmitted. Should not be higher than 127. Default value
648 is 6, which corresponds to 63seconds till the last retransmission
649 with the current initial RTO of 1second. With this the final timeout
650 for an active TCP connection attempt will happen after 127seconds.
652 tcp_timestamps - INTEGER
653 Enable timestamps as defined in RFC1323.
655 1: Enable timestamps as defined in RFC1323 and use random offset for
656 each connection rather than only using the current time.
657 2: Like 1, but without random offsets.
660 tcp_min_tso_segs - INTEGER
661 Minimal number of segments per TSO frame.
662 Since linux-3.12, TCP does an automatic sizing of TSO frames,
663 depending on flow rate, instead of filling 64Kbytes packets.
664 For specific usages, it's possible to force TCP to build big
665 TSO frames. Note that TCP stack might split too big TSO packets
666 if available window is too small.
669 tcp_pacing_ss_ratio - INTEGER
670 sk->sk_pacing_rate is set by TCP stack using a ratio applied
671 to current rate. (current_rate = cwnd * mss / srtt)
672 If TCP is in slow start, tcp_pacing_ss_ratio is applied
673 to let TCP probe for bigger speeds, assuming cwnd can be
674 doubled every other RTT.
677 tcp_pacing_ca_ratio - INTEGER
678 sk->sk_pacing_rate is set by TCP stack using a ratio applied
679 to current rate. (current_rate = cwnd * mss / srtt)
680 If TCP is in congestion avoidance phase, tcp_pacing_ca_ratio
681 is applied to conservatively probe for bigger throughput.
684 tcp_tso_win_divisor - INTEGER
685 This allows control over what percentage of the congestion window
686 can be consumed by a single TSO frame.
687 The setting of this parameter is a choice between burstiness and
688 building larger TSO frames.
691 tcp_tw_reuse - INTEGER
692 Enable reuse of TIME-WAIT sockets for new connections when it is
693 safe from protocol viewpoint.
696 2 - enable for loopback traffic only
697 It should not be changed without advice/request of technical
701 tcp_window_scaling - BOOLEAN
702 Enable window scaling as defined in RFC1323.
704 tcp_wmem - vector of 3 INTEGERs: min, default, max
705 min: Amount of memory reserved for send buffers for TCP sockets.
706 Each TCP socket has rights to use it due to fact of its birth.
709 default: initial size of send buffer used by TCP sockets. This
710 value overrides net.core.wmem_default used by other protocols.
711 It is usually lower than net.core.wmem_default.
714 max: Maximal amount of memory allowed for automatically tuned
715 send buffers for TCP sockets. This value does not override
716 net.core.wmem_max. Calling setsockopt() with SO_SNDBUF disables
717 automatic tuning of that socket's send buffer size, in which case
718 this value is ignored.
719 Default: between 64K and 4MB, depending on RAM size.
721 tcp_notsent_lowat - UNSIGNED INTEGER
722 A TCP socket can control the amount of unsent bytes in its write queue,
723 thanks to TCP_NOTSENT_LOWAT socket option. poll()/select()/epoll()
724 reports POLLOUT events if the amount of unsent bytes is below a per
725 socket value, and if the write queue is not full. sendmsg() will
726 also not add new buffers if the limit is hit.
728 This global variable controls the amount of unsent data for
729 sockets not using TCP_NOTSENT_LOWAT. For these sockets, a change
730 to the global variable has immediate effect.
732 Default: UINT_MAX (0xFFFFFFFF)
734 tcp_workaround_signed_windows - BOOLEAN
735 If set, assume no receipt of a window scaling option means the
736 remote TCP is broken and treats the window as a signed quantity.
737 If unset, assume the remote TCP is not broken even if we do
738 not receive a window scaling option from them.
741 tcp_thin_linear_timeouts - BOOLEAN
742 Enable dynamic triggering of linear timeouts for thin streams.
743 If set, a check is performed upon retransmission by timeout to
744 determine if the stream is thin (less than 4 packets in flight).
745 As long as the stream is found to be thin, up to 6 linear
746 timeouts may be performed before exponential backoff mode is
747 initiated. This improves retransmission latency for
748 non-aggressive thin streams, often found to be time-dependent.
749 For more information on thin streams, see
750 Documentation/networking/tcp-thin.txt
753 tcp_limit_output_bytes - INTEGER
754 Controls TCP Small Queue limit per tcp socket.
755 TCP bulk sender tends to increase packets in flight until it
756 gets losses notifications. With SNDBUF autotuning, this can
757 result in a large amount of packets queued on the local machine
758 (e.g.: qdiscs, CPU backlog, or device) hurting latency of other
759 flows, for typical pfifo_fast qdiscs. tcp_limit_output_bytes
760 limits the number of bytes on qdisc or device to reduce artificial
761 RTT/cwnd and reduce bufferbloat.
762 Default: 1048576 (16 * 65536)
764 tcp_challenge_ack_limit - INTEGER
765 Limits number of Challenge ACK sent per second, as recommended
766 in RFC 5961 (Improving TCP's Robustness to Blind In-Window Attacks)
771 udp_l3mdev_accept - BOOLEAN
772 Enabling this option allows a "global" bound socket to work
773 across L3 master domains (e.g., VRFs) with packets capable of
774 being received regardless of the L3 domain in which they
775 originated. Only valid when the kernel was compiled with
776 CONFIG_NET_L3_MASTER_DEV.
777 Default: 0 (disabled)
779 udp_mem - vector of 3 INTEGERs: min, pressure, max
780 Number of pages allowed for queueing by all UDP sockets.
782 min: Below this number of pages UDP is not bothered about its
783 memory appetite. When amount of memory allocated by UDP exceeds
784 this number, UDP starts to moderate memory usage.
786 pressure: This value was introduced to follow format of tcp_mem.
788 max: Number of pages allowed for queueing by all UDP sockets.
790 Default is calculated at boot time from amount of available memory.
792 udp_rmem_min - INTEGER
793 Minimal size of receive buffer used by UDP sockets in moderation.
794 Each UDP socket is able to use the size for receiving data, even if
795 total pages of UDP sockets exceed udp_mem pressure. The unit is byte.
798 udp_wmem_min - INTEGER
799 Minimal size of send buffer used by UDP sockets in moderation.
800 Each UDP socket is able to use the size for sending data, even if
801 total pages of UDP sockets exceed udp_mem pressure. The unit is byte.
806 raw_l3mdev_accept - BOOLEAN
807 Enabling this option allows a "global" bound socket to work
808 across L3 master domains (e.g., VRFs) with packets capable of
809 being received regardless of the L3 domain in which they
810 originated. Only valid when the kernel was compiled with
811 CONFIG_NET_L3_MASTER_DEV.
816 cipso_cache_enable - BOOLEAN
817 If set, enable additions to and lookups from the CIPSO label mapping
818 cache. If unset, additions are ignored and lookups always result in a
819 miss. However, regardless of the setting the cache is still
820 invalidated when required when means you can safely toggle this on and
821 off and the cache will always be "safe".
824 cipso_cache_bucket_size - INTEGER
825 The CIPSO label cache consists of a fixed size hash table with each
826 hash bucket containing a number of cache entries. This variable limits
827 the number of entries in each hash bucket; the larger the value the
828 more CIPSO label mappings that can be cached. When the number of
829 entries in a given hash bucket reaches this limit adding new entries
830 causes the oldest entry in the bucket to be removed to make room.
833 cipso_rbm_optfmt - BOOLEAN
834 Enable the "Optimized Tag 1 Format" as defined in section 3.4.2.6 of
835 the CIPSO draft specification (see Documentation/netlabel for details).
836 This means that when set the CIPSO tag will be padded with empty
837 categories in order to make the packet data 32-bit aligned.
840 cipso_rbm_structvalid - BOOLEAN
841 If set, do a very strict check of the CIPSO option when
842 ip_options_compile() is called. If unset, relax the checks done during
843 ip_options_compile(). Either way is "safe" as errors are caught else
844 where in the CIPSO processing code but setting this to 0 (False) should
845 result in less work (i.e. it should be faster) but could cause problems
846 with other implementations that require strict checking.
851 ip_local_port_range - 2 INTEGERS
852 Defines the local port range that is used by TCP and UDP to
853 choose the local port. The first number is the first, the
854 second the last local port number.
855 If possible, it is better these numbers have different parity.
856 (one even and one odd values)
857 The default values are 32768 and 60999 respectively.
859 ip_local_reserved_ports - list of comma separated ranges
860 Specify the ports which are reserved for known third-party
861 applications. These ports will not be used by automatic port
862 assignments (e.g. when calling connect() or bind() with port
863 number 0). Explicit port allocation behavior is unchanged.
865 The format used for both input and output is a comma separated
866 list of ranges (e.g. "1,2-4,10-10" for ports 1, 2, 3, 4 and
867 10). Writing to the file will clear all previously reserved
868 ports and update the current list with the one given in the
871 Note that ip_local_port_range and ip_local_reserved_ports
872 settings are independent and both are considered by the kernel
873 when determining which ports are available for automatic port
876 You can reserve ports which are not in the current
877 ip_local_port_range, e.g.:
879 $ cat /proc/sys/net/ipv4/ip_local_port_range
881 $ cat /proc/sys/net/ipv4/ip_local_reserved_ports
884 although this is redundant. However such a setting is useful
885 if later the port range is changed to a value that will
886 include the reserved ports.
890 ip_unprivileged_port_start - INTEGER
891 This is a per-namespace sysctl. It defines the first
892 unprivileged port in the network namespace. Privileged ports
893 require root or CAP_NET_BIND_SERVICE in order to bind to them.
894 To disable all privileged ports, set this to 0. It may not
895 overlap with the ip_local_reserved_ports range.
899 ip_nonlocal_bind - BOOLEAN
900 If set, allows processes to bind() to non-local IP addresses,
901 which can be quite useful - but may break some applications.
905 If set non-zero, enables support for dynamic addresses.
906 If set to a non-zero value larger than 1, a kernel log
907 message will be printed when dynamic address rewriting
911 ip_early_demux - BOOLEAN
912 Optimize input packet processing down to one demux for
913 certain kinds of local sockets. Currently we only do this
914 for established TCP and connected UDP sockets.
916 It may add an additional cost for pure routing workloads that
917 reduces overall throughput, in such case you should disable it.
920 tcp_early_demux - BOOLEAN
921 Enable early demux for established TCP sockets.
924 udp_early_demux - BOOLEAN
925 Enable early demux for connected UDP sockets. Disable this if
926 your system could experience more unconnected load.
929 icmp_echo_ignore_all - BOOLEAN
930 If set non-zero, then the kernel will ignore all ICMP ECHO
934 icmp_echo_ignore_broadcasts - BOOLEAN
935 If set non-zero, then the kernel will ignore all ICMP ECHO and
936 TIMESTAMP requests sent to it via broadcast/multicast.
939 icmp_ratelimit - INTEGER
940 Limit the maximal rates for sending ICMP packets whose type matches
941 icmp_ratemask (see below) to specific targets.
942 0 to disable any limiting,
943 otherwise the minimal space between responses in milliseconds.
944 Note that another sysctl, icmp_msgs_per_sec limits the number
945 of ICMP packets sent on all targets.
948 icmp_msgs_per_sec - INTEGER
949 Limit maximal number of ICMP packets sent per second from this host.
950 Only messages whose type matches icmp_ratemask (see below) are
951 controlled by this limit.
954 icmp_msgs_burst - INTEGER
955 icmp_msgs_per_sec controls number of ICMP packets sent per second,
956 while icmp_msgs_burst controls the burst size of these packets.
959 icmp_ratemask - INTEGER
960 Mask made of ICMP types for which rates are being limited.
961 Significant bits: IHGFEDCBA9876543210
962 Default mask: 0000001100000011000 (6168)
964 Bit definitions (see include/linux/icmp.h):
966 3 Destination Unreachable *
971 C Parameter Problem *
976 H Address Mask Request
979 * These are rate limited by default (see default mask above)
981 icmp_ignore_bogus_error_responses - BOOLEAN
982 Some routers violate RFC1122 by sending bogus responses to broadcast
983 frames. Such violations are normally logged via a kernel warning.
984 If this is set to TRUE, the kernel will not give such warnings, which
985 will avoid log file clutter.
988 icmp_errors_use_inbound_ifaddr - BOOLEAN
990 If zero, icmp error messages are sent with the primary address of
991 the exiting interface.
993 If non-zero, the message will be sent with the primary address of
994 the interface that received the packet that caused the icmp error.
995 This is the behaviour network many administrators will expect from
996 a router. And it can make debugging complicated network layouts
999 Note that if no primary address exists for the interface selected,
1000 then the primary address of the first non-loopback interface that
1001 has one will be used regardless of this setting.
1005 igmp_max_memberships - INTEGER
1006 Change the maximum number of multicast groups we can subscribe to.
1009 Theoretical maximum value is bounded by having to send a membership
1010 report in a single datagram (i.e. the report can't span multiple
1011 datagrams, or risk confusing the switch and leaving groups you don't
1014 The number of supported groups 'M' is bounded by the number of group
1015 report entries you can fit into a single datagram of 65535 bytes.
1017 M = 65536-sizeof (ip header)/(sizeof(Group record))
1019 Group records are variable length, with a minimum of 12 bytes.
1020 So net.ipv4.igmp_max_memberships should not be set higher than:
1022 (65536-24) / 12 = 5459
1024 The value 5459 assumes no IP header options, so in practice
1025 this number may be lower.
1027 igmp_max_msf - INTEGER
1028 Maximum number of addresses allowed in the source filter list for a
1033 Controls the IGMP query robustness variable (see RFC2236 8.1).
1034 Default: 2 (as specified by RFC2236 8.1)
1035 Minimum: 1 (as specified by RFC6636 4.5)
1037 force_igmp_version - INTEGER
1038 0 - (default) No enforcement of a IGMP version, IGMPv1/v2 fallback
1039 allowed. Will back to IGMPv3 mode again if all IGMPv1/v2 Querier
1040 Present timer expires.
1041 1 - Enforce to use IGMP version 1. Will also reply IGMPv1 report if
1042 receive IGMPv2/v3 query.
1043 2 - Enforce to use IGMP version 2. Will fallback to IGMPv1 if receive
1044 IGMPv1 query message. Will reply report if receive IGMPv3 query.
1045 3 - Enforce to use IGMP version 3. The same react with default 0.
1047 Note: this is not the same with force_mld_version because IGMPv3 RFC3376
1048 Security Considerations does not have clear description that we could
1049 ignore other version messages completely as MLDv2 RFC3810. So make
1050 this value as default 0 is recommended.
1052 conf/interface/* changes special settings per interface (where
1053 "interface" is the name of your network interface)
1055 conf/all/* is special, changes the settings for all interfaces
1057 log_martians - BOOLEAN
1058 Log packets with impossible addresses to kernel log.
1059 log_martians for the interface will be enabled if at least one of
1060 conf/{all,interface}/log_martians is set to TRUE,
1061 it will be disabled otherwise
1063 accept_redirects - BOOLEAN
1064 Accept ICMP redirect messages.
1065 accept_redirects for the interface will be enabled if:
1066 - both conf/{all,interface}/accept_redirects are TRUE in the case
1067 forwarding for the interface is enabled
1069 - at least one of conf/{all,interface}/accept_redirects is TRUE in the
1070 case forwarding for the interface is disabled
1071 accept_redirects for the interface will be disabled otherwise
1075 forwarding - BOOLEAN
1076 Enable IP forwarding on this interface. This controls whether packets
1077 received _on_ this interface can be forwarded.
1079 mc_forwarding - BOOLEAN
1080 Do multicast routing. The kernel needs to be compiled with CONFIG_MROUTE
1081 and a multicast routing daemon is required.
1082 conf/all/mc_forwarding must also be set to TRUE to enable multicast
1083 routing for the interface
1086 Integer value used to differentiate the devices by the medium they
1087 are attached to. Two devices can have different id values when
1088 the broadcast packets are received only on one of them.
1089 The default value 0 means that the device is the only interface
1090 to its medium, value of -1 means that medium is not known.
1092 Currently, it is used to change the proxy_arp behavior:
1093 the proxy_arp feature is enabled for packets forwarded between
1094 two devices attached to different media.
1098 proxy_arp for the interface will be enabled if at least one of
1099 conf/{all,interface}/proxy_arp is set to TRUE,
1100 it will be disabled otherwise
1102 proxy_arp_pvlan - BOOLEAN
1103 Private VLAN proxy arp.
1104 Basically allow proxy arp replies back to the same interface
1105 (from which the ARP request/solicitation was received).
1107 This is done to support (ethernet) switch features, like RFC
1108 3069, where the individual ports are NOT allowed to
1109 communicate with each other, but they are allowed to talk to
1110 the upstream router. As described in RFC 3069, it is possible
1111 to allow these hosts to communicate through the upstream
1112 router by proxy_arp'ing. Don't need to be used together with
1115 This technology is known by different names:
1116 In RFC 3069 it is called VLAN Aggregation.
1117 Cisco and Allied Telesyn call it Private VLAN.
1118 Hewlett-Packard call it Source-Port filtering or port-isolation.
1119 Ericsson call it MAC-Forced Forwarding (RFC Draft).
1121 shared_media - BOOLEAN
1122 Send(router) or accept(host) RFC1620 shared media redirects.
1123 Overrides secure_redirects.
1124 shared_media for the interface will be enabled if at least one of
1125 conf/{all,interface}/shared_media is set to TRUE,
1126 it will be disabled otherwise
1129 secure_redirects - BOOLEAN
1130 Accept ICMP redirect messages only to gateways listed in the
1131 interface's current gateway list. Even if disabled, RFC1122 redirect
1133 Overridden by shared_media.
1134 secure_redirects for the interface will be enabled if at least one of
1135 conf/{all,interface}/secure_redirects is set to TRUE,
1136 it will be disabled otherwise
1139 send_redirects - BOOLEAN
1140 Send redirects, if router.
1141 send_redirects for the interface will be enabled if at least one of
1142 conf/{all,interface}/send_redirects is set to TRUE,
1143 it will be disabled otherwise
1146 bootp_relay - BOOLEAN
1147 Accept packets with source address 0.b.c.d destined
1148 not to this host as local ones. It is supposed, that
1149 BOOTP relay daemon will catch and forward such packets.
1150 conf/all/bootp_relay must also be set to TRUE to enable BOOTP relay
1153 Not Implemented Yet.
1155 accept_source_route - BOOLEAN
1156 Accept packets with SRR option.
1157 conf/all/accept_source_route must also be set to TRUE to accept packets
1158 with SRR option on the interface
1159 default TRUE (router)
1162 accept_local - BOOLEAN
1163 Accept packets with local source addresses. In combination with
1164 suitable routing, this can be used to direct packets between two
1165 local interfaces over the wire and have them accepted properly.
1168 route_localnet - BOOLEAN
1169 Do not consider loopback addresses as martian source or destination
1170 while routing. This enables the use of 127/8 for local routing purposes.
1174 0 - No source validation.
1175 1 - Strict mode as defined in RFC3704 Strict Reverse Path
1176 Each incoming packet is tested against the FIB and if the interface
1177 is not the best reverse path the packet check will fail.
1178 By default failed packets are discarded.
1179 2 - Loose mode as defined in RFC3704 Loose Reverse Path
1180 Each incoming packet's source address is also tested against the FIB
1181 and if the source address is not reachable via any interface
1182 the packet check will fail.
1184 Current recommended practice in RFC3704 is to enable strict mode
1185 to prevent IP spoofing from DDos attacks. If using asymmetric routing
1186 or other complicated routing, then loose mode is recommended.
1188 The max value from conf/{all,interface}/rp_filter is used
1189 when doing source validation on the {interface}.
1191 Default value is 0. Note that some distributions enable it
1194 arp_filter - BOOLEAN
1195 1 - Allows you to have multiple network interfaces on the same
1196 subnet, and have the ARPs for each interface be answered
1197 based on whether or not the kernel would route a packet from
1198 the ARP'd IP out that interface (therefore you must use source
1199 based routing for this to work). In other words it allows control
1200 of which cards (usually 1) will respond to an arp request.
1202 0 - (default) The kernel can respond to arp requests with addresses
1203 from other interfaces. This may seem wrong but it usually makes
1204 sense, because it increases the chance of successful communication.
1205 IP addresses are owned by the complete host on Linux, not by
1206 particular interfaces. Only for more complex setups like load-
1207 balancing, does this behaviour cause problems.
1209 arp_filter for the interface will be enabled if at least one of
1210 conf/{all,interface}/arp_filter is set to TRUE,
1211 it will be disabled otherwise
1213 arp_announce - INTEGER
1214 Define different restriction levels for announcing the local
1215 source IP address from IP packets in ARP requests sent on
1217 0 - (default) Use any local address, configured on any interface
1218 1 - Try to avoid local addresses that are not in the target's
1219 subnet for this interface. This mode is useful when target
1220 hosts reachable via this interface require the source IP
1221 address in ARP requests to be part of their logical network
1222 configured on the receiving interface. When we generate the
1223 request we will check all our subnets that include the
1224 target IP and will preserve the source address if it is from
1225 such subnet. If there is no such subnet we select source
1226 address according to the rules for level 2.
1227 2 - Always use the best local address for this target.
1228 In this mode we ignore the source address in the IP packet
1229 and try to select local address that we prefer for talks with
1230 the target host. Such local address is selected by looking
1231 for primary IP addresses on all our subnets on the outgoing
1232 interface that include the target IP address. If no suitable
1233 local address is found we select the first local address
1234 we have on the outgoing interface or on all other interfaces,
1235 with the hope we will receive reply for our request and
1236 even sometimes no matter the source IP address we announce.
1238 The max value from conf/{all,interface}/arp_announce is used.
1240 Increasing the restriction level gives more chance for
1241 receiving answer from the resolved target while decreasing
1242 the level announces more valid sender's information.
1244 arp_ignore - INTEGER
1245 Define different modes for sending replies in response to
1246 received ARP requests that resolve local target IP addresses:
1247 0 - (default): reply for any local target IP address, configured
1249 1 - reply only if the target IP address is local address
1250 configured on the incoming interface
1251 2 - reply only if the target IP address is local address
1252 configured on the incoming interface and both with the
1253 sender's IP address are part from same subnet on this interface
1254 3 - do not reply for local addresses configured with scope host,
1255 only resolutions for global and link addresses are replied
1257 8 - do not reply for all local addresses
1259 The max value from conf/{all,interface}/arp_ignore is used
1260 when ARP request is received on the {interface}
1262 arp_notify - BOOLEAN
1263 Define mode for notification of address and device changes.
1264 0 - (default): do nothing
1265 1 - Generate gratuitous arp requests when device is brought up
1266 or hardware address changes.
1268 arp_accept - BOOLEAN
1269 Define behavior for gratuitous ARP frames who's IP is not
1270 already present in the ARP table:
1271 0 - don't create new entries in the ARP table
1272 1 - create new entries in the ARP table
1274 Both replies and requests type gratuitous arp will trigger the
1275 ARP table to be updated, if this setting is on.
1277 If the ARP table already contains the IP address of the
1278 gratuitous arp frame, the arp table will be updated regardless
1279 if this setting is on or off.
1281 mcast_solicit - INTEGER
1282 The maximum number of multicast probes in INCOMPLETE state,
1283 when the associated hardware address is unknown. Defaults
1286 ucast_solicit - INTEGER
1287 The maximum number of unicast probes in PROBE state, when
1288 the hardware address is being reconfirmed. Defaults to 3.
1290 app_solicit - INTEGER
1291 The maximum number of probes to send to the user space ARP daemon
1292 via netlink before dropping back to multicast probes (see
1293 mcast_resolicit). Defaults to 0.
1295 mcast_resolicit - INTEGER
1296 The maximum number of multicast probes after unicast and
1297 app probes in PROBE state. Defaults to 0.
1299 disable_policy - BOOLEAN
1300 Disable IPSEC policy (SPD) for this interface
1302 disable_xfrm - BOOLEAN
1303 Disable IPSEC encryption on this interface, whatever the policy
1305 igmpv2_unsolicited_report_interval - INTEGER
1306 The interval in milliseconds in which the next unsolicited
1307 IGMPv1 or IGMPv2 report retransmit will take place.
1308 Default: 10000 (10 seconds)
1310 igmpv3_unsolicited_report_interval - INTEGER
1311 The interval in milliseconds in which the next unsolicited
1312 IGMPv3 report retransmit will take place.
1313 Default: 1000 (1 seconds)
1315 promote_secondaries - BOOLEAN
1316 When a primary IP address is removed from this interface
1317 promote a corresponding secondary IP address instead of
1318 removing all the corresponding secondary IP addresses.
1320 drop_unicast_in_l2_multicast - BOOLEAN
1321 Drop any unicast IP packets that are received in link-layer
1322 multicast (or broadcast) frames.
1323 This behavior (for multicast) is actually a SHOULD in RFC
1324 1122, but is disabled by default for compatibility reasons.
1327 drop_gratuitous_arp - BOOLEAN
1328 Drop all gratuitous ARP frames, for example if there's a known
1329 good ARP proxy on the network and such frames need not be used
1330 (or in the case of 802.11, must not be used to prevent attacks.)
1335 Allows you to write a number, which can be used as required.
1338 xfrm4_gc_thresh - INTEGER
1339 The threshold at which we will start garbage collecting for IPv4
1340 destination cache entries. At twice this value the system will
1341 refuse new allocations.
1343 igmp_link_local_mcast_reports - BOOLEAN
1344 Enable IGMP reports for link local multicast groups in the
1349 kuznet@ms2.inr.ac.ru
1355 delon.nicolas@wanadoo.fr
1360 /proc/sys/net/ipv6/* Variables:
1362 IPv6 has no global variables such as tcp_*. tcp_* settings under ipv4/ also
1363 apply to IPv6 [XXX?].
1365 bindv6only - BOOLEAN
1366 Default value for IPV6_V6ONLY socket option,
1367 which restricts use of the IPv6 socket to IPv6 communication
1369 TRUE: disable IPv4-mapped address feature
1370 FALSE: enable IPv4-mapped address feature
1372 Default: FALSE (as specified in RFC3493)
1374 flowlabel_consistency - BOOLEAN
1375 Protect the consistency (and unicity) of flow label.
1376 You have to disable it to use IPV6_FL_F_REFLECT flag on the
1382 auto_flowlabels - INTEGER
1383 Automatically generate flow labels based on a flow hash of the
1384 packet. This allows intermediate devices, such as routers, to
1385 identify packet flows for mechanisms like Equal Cost Multipath
1386 Routing (see RFC 6438).
1387 0: automatic flow labels are completely disabled
1388 1: automatic flow labels are enabled by default, they can be
1389 disabled on a per socket basis using the IPV6_AUTOFLOWLABEL
1391 2: automatic flow labels are allowed, they may be enabled on a
1392 per socket basis using the IPV6_AUTOFLOWLABEL socket option
1393 3: automatic flow labels are enabled and enforced, they cannot
1394 be disabled by the socket option
1397 flowlabel_state_ranges - BOOLEAN
1398 Split the flow label number space into two ranges. 0-0x7FFFF is
1399 reserved for the IPv6 flow manager facility, 0x80000-0xFFFFF
1400 is reserved for stateless flow labels as described in RFC6437.
1405 flowlabel_reflect - BOOLEAN
1406 Automatically reflect the flow label. Needed for Path MTU
1407 Discovery to work with Equal Cost Multipath Routing in anycast
1408 environments. See RFC 7690 and:
1409 https://tools.ietf.org/html/draft-wang-6man-flow-label-reflection-01
1414 fib_multipath_hash_policy - INTEGER
1415 Controls which hash policy to use for multipath routes.
1416 Default: 0 (Layer 3)
1418 0 - Layer 3 (source and destination addresses plus flow label)
1419 1 - Layer 4 (standard 5-tuple)
1421 anycast_src_echo_reply - BOOLEAN
1422 Controls the use of anycast addresses as source addresses for ICMPv6
1428 idgen_delay - INTEGER
1429 Controls the delay in seconds after which time to retry
1430 privacy stable address generation if a DAD conflict is
1432 Default: 1 (as specified in RFC7217)
1434 idgen_retries - INTEGER
1435 Controls the number of retries to generate a stable privacy
1436 address if a DAD conflict is detected.
1437 Default: 3 (as specified in RFC7217)
1440 Controls the MLD query robustness variable (see RFC3810 9.1).
1441 Default: 2 (as specified by RFC3810 9.1)
1442 Minimum: 1 (as specified by RFC6636 4.5)
1444 max_dst_opts_number - INTEGER
1445 Maximum number of non-padding TLVs allowed in a Destination
1446 options extension header. If this value is less than zero
1447 then unknown options are disallowed and the number of known
1448 TLVs allowed is the absolute value of this number.
1451 max_hbh_opts_number - INTEGER
1452 Maximum number of non-padding TLVs allowed in a Hop-by-Hop
1453 options extension header. If this value is less than zero
1454 then unknown options are disallowed and the number of known
1455 TLVs allowed is the absolute value of this number.
1458 max_dst_opts_length - INTEGER
1459 Maximum length allowed for a Destination options extension
1461 Default: INT_MAX (unlimited)
1463 max_hbh_length - INTEGER
1464 Maximum length allowed for a Hop-by-Hop options extension
1466 Default: INT_MAX (unlimited)
1468 skip_notify_on_dev_down - BOOLEAN
1469 Controls whether an RTM_DELROUTE message is generated for routes
1470 removed when a device is taken down or deleted. IPv4 does not
1471 generate this message; IPv6 does by default. Setting this sysctl
1472 to true skips the message, making IPv4 and IPv6 on par in relying
1473 on userspace caches to track link events and evict routes.
1474 Default: false (generate message)
1478 ip6frag_high_thresh - INTEGER
1479 Maximum memory used to reassemble IPv6 fragments. When
1480 ip6frag_high_thresh bytes of memory is allocated for this purpose,
1481 the fragment handler will toss packets until ip6frag_low_thresh
1484 ip6frag_low_thresh - INTEGER
1485 See ip6frag_high_thresh
1487 ip6frag_time - INTEGER
1488 Time in seconds to keep an IPv6 fragment in memory.
1490 IPv6 Segment Routing:
1492 seg6_flowlabel - INTEGER
1493 Controls the behaviour of computing the flowlabel of outer
1494 IPv6 header in case of SR T.encaps
1496 -1 set flowlabel to zero.
1497 0 copy flowlabel from Inner packet in case of Inner IPv6
1498 (Set flowlabel to 0 in case IPv4/L2)
1499 1 Compute the flowlabel using seg6_make_flowlabel()
1504 Change the interface-specific default settings.
1508 Change all the interface-specific settings.
1510 [XXX: Other special features than forwarding?]
1512 conf/all/forwarding - BOOLEAN
1513 Enable global IPv6 forwarding between all interfaces.
1515 IPv4 and IPv6 work differently here; e.g. netfilter must be used
1516 to control which interfaces may forward packets and which not.
1518 This also sets all interfaces' Host/Router setting
1519 'forwarding' to the specified value. See below for details.
1521 This referred to as global forwarding.
1526 fwmark_reflect - BOOLEAN
1527 Controls the fwmark of kernel-generated IPv6 reply packets that are not
1528 associated with a socket for example, TCP RSTs or ICMPv6 echo replies).
1529 If unset, these packets have a fwmark of zero. If set, they have the
1530 fwmark of the packet they are replying to.
1534 Change special settings per interface.
1536 The functional behaviour for certain settings is different
1537 depending on whether local forwarding is enabled or not.
1540 Accept Router Advertisements; autoconfigure using them.
1542 It also determines whether or not to transmit Router
1543 Solicitations. If and only if the functional setting is to
1544 accept Router Advertisements, Router Solicitations will be
1547 Possible values are:
1548 0 Do not accept Router Advertisements.
1549 1 Accept Router Advertisements if forwarding is disabled.
1550 2 Overrule forwarding behaviour. Accept Router Advertisements
1551 even if forwarding is enabled.
1553 Functional default: enabled if local forwarding is disabled.
1554 disabled if local forwarding is enabled.
1556 accept_ra_defrtr - BOOLEAN
1557 Learn default router in Router Advertisement.
1559 Functional default: enabled if accept_ra is enabled.
1560 disabled if accept_ra is disabled.
1562 accept_ra_from_local - BOOLEAN
1563 Accept RA with source-address that is found on local machine
1564 if the RA is otherwise proper and able to be accepted.
1565 Default is to NOT accept these as it may be an un-intended
1569 enabled if accept_ra_from_local is enabled
1570 on a specific interface.
1571 disabled if accept_ra_from_local is disabled
1572 on a specific interface.
1574 accept_ra_min_hop_limit - INTEGER
1575 Minimum hop limit Information in Router Advertisement.
1577 Hop limit Information in Router Advertisement less than this
1578 variable shall be ignored.
1582 accept_ra_pinfo - BOOLEAN
1583 Learn Prefix Information in Router Advertisement.
1585 Functional default: enabled if accept_ra is enabled.
1586 disabled if accept_ra is disabled.
1588 accept_ra_rt_info_min_plen - INTEGER
1589 Minimum prefix length of Route Information in RA.
1591 Route Information w/ prefix smaller than this variable shall
1594 Functional default: 0 if accept_ra_rtr_pref is enabled.
1595 -1 if accept_ra_rtr_pref is disabled.
1597 accept_ra_rt_info_max_plen - INTEGER
1598 Maximum prefix length of Route Information in RA.
1600 Route Information w/ prefix larger than this variable shall
1603 Functional default: 0 if accept_ra_rtr_pref is enabled.
1604 -1 if accept_ra_rtr_pref is disabled.
1606 accept_ra_rtr_pref - BOOLEAN
1607 Accept Router Preference in RA.
1609 Functional default: enabled if accept_ra is enabled.
1610 disabled if accept_ra is disabled.
1612 accept_ra_mtu - BOOLEAN
1613 Apply the MTU value specified in RA option 5 (RFC4861). If
1614 disabled, the MTU specified in the RA will be ignored.
1616 Functional default: enabled if accept_ra is enabled.
1617 disabled if accept_ra is disabled.
1619 accept_redirects - BOOLEAN
1622 Functional default: enabled if local forwarding is disabled.
1623 disabled if local forwarding is enabled.
1625 accept_source_route - INTEGER
1626 Accept source routing (routing extension header).
1628 >= 0: Accept only routing header type 2.
1629 < 0: Do not accept routing header.
1634 Autoconfigure addresses using Prefix Information in Router
1637 Functional default: enabled if accept_ra_pinfo is enabled.
1638 disabled if accept_ra_pinfo is disabled.
1640 dad_transmits - INTEGER
1641 The amount of Duplicate Address Detection probes to send.
1644 forwarding - INTEGER
1645 Configure interface-specific Host/Router behaviour.
1647 Note: It is recommended to have the same setting on all
1648 interfaces; mixed router/host scenarios are rather uncommon.
1650 Possible values are:
1651 0 Forwarding disabled
1652 1 Forwarding enabled
1656 By default, Host behaviour is assumed. This means:
1658 1. IsRouter flag is not set in Neighbour Advertisements.
1659 2. If accept_ra is TRUE (default), transmit Router
1661 3. If accept_ra is TRUE (default), accept Router
1662 Advertisements (and do autoconfiguration).
1663 4. If accept_redirects is TRUE (default), accept Redirects.
1667 If local forwarding is enabled, Router behaviour is assumed.
1668 This means exactly the reverse from the above:
1670 1. IsRouter flag is set in Neighbour Advertisements.
1671 2. Router Solicitations are not sent unless accept_ra is 2.
1672 3. Router Advertisements are ignored unless accept_ra is 2.
1673 4. Redirects are ignored.
1675 Default: 0 (disabled) if global forwarding is disabled (default),
1676 otherwise 1 (enabled).
1679 Default Hop Limit to set.
1683 Default Maximum Transfer Unit
1684 Default: 1280 (IPv6 required minimum)
1686 ip_nonlocal_bind - BOOLEAN
1687 If set, allows processes to bind() to non-local IPv6 addresses,
1688 which can be quite useful - but may break some applications.
1691 router_probe_interval - INTEGER
1692 Minimum interval (in seconds) between Router Probing described
1697 router_solicitation_delay - INTEGER
1698 Number of seconds to wait after interface is brought up
1699 before sending Router Solicitations.
1702 router_solicitation_interval - INTEGER
1703 Number of seconds to wait between Router Solicitations.
1706 router_solicitations - INTEGER
1707 Number of Router Solicitations to send until assuming no
1708 routers are present.
1711 use_oif_addrs_only - BOOLEAN
1712 When enabled, the candidate source addresses for destinations
1713 routed via this interface are restricted to the set of addresses
1714 configured on this interface (vis. RFC 6724, section 4).
1718 use_tempaddr - INTEGER
1719 Preference for Privacy Extensions (RFC3041).
1720 <= 0 : disable Privacy Extensions
1721 == 1 : enable Privacy Extensions, but prefer public
1722 addresses over temporary addresses.
1723 > 1 : enable Privacy Extensions and prefer temporary
1724 addresses over public addresses.
1725 Default: 0 (for most devices)
1726 -1 (for point-to-point devices and loopback devices)
1728 temp_valid_lft - INTEGER
1729 valid lifetime (in seconds) for temporary addresses.
1730 Default: 604800 (7 days)
1732 temp_prefered_lft - INTEGER
1733 Preferred lifetime (in seconds) for temporary addresses.
1734 Default: 86400 (1 day)
1736 keep_addr_on_down - INTEGER
1737 Keep all IPv6 addresses on an interface down event. If set static
1738 global addresses with no expiration time are not flushed.
1743 Default: 0 (addresses are removed)
1745 max_desync_factor - INTEGER
1746 Maximum value for DESYNC_FACTOR, which is a random value
1747 that ensures that clients don't synchronize with each
1748 other and generate new addresses at exactly the same time.
1749 value is in seconds.
1752 regen_max_retry - INTEGER
1753 Number of attempts before give up attempting to generate
1754 valid temporary addresses.
1757 max_addresses - INTEGER
1758 Maximum number of autoconfigured addresses per interface. Setting
1759 to zero disables the limitation. It is not recommended to set this
1760 value too large (or to zero) because it would be an easy way to
1761 crash the kernel by allowing too many addresses to be created.
1764 disable_ipv6 - BOOLEAN
1765 Disable IPv6 operation. If accept_dad is set to 2, this value
1766 will be dynamically set to TRUE if DAD fails for the link-local
1768 Default: FALSE (enable IPv6 operation)
1770 When this value is changed from 1 to 0 (IPv6 is being enabled),
1771 it will dynamically create a link-local address on the given
1772 interface and start Duplicate Address Detection, if necessary.
1774 When this value is changed from 0 to 1 (IPv6 is being disabled),
1775 it will dynamically delete all addresses and routes on the given
1776 interface. From now on it will not possible to add addresses/routes
1777 to the selected interface.
1779 accept_dad - INTEGER
1780 Whether to accept DAD (Duplicate Address Detection).
1782 1: Enable DAD (default)
1783 2: Enable DAD, and disable IPv6 operation if MAC-based duplicate
1784 link-local address has been found.
1786 DAD operation and mode on a given interface will be selected according
1787 to the maximum value of conf/{all,interface}/accept_dad.
1789 force_tllao - BOOLEAN
1790 Enable sending the target link-layer address option even when
1791 responding to a unicast neighbor solicitation.
1794 Quoting from RFC 2461, section 4.4, Target link-layer address:
1796 "The option MUST be included for multicast solicitations in order to
1797 avoid infinite Neighbor Solicitation "recursion" when the peer node
1798 does not have a cache entry to return a Neighbor Advertisements
1799 message. When responding to unicast solicitations, the option can be
1800 omitted since the sender of the solicitation has the correct link-
1801 layer address; otherwise it would not have be able to send the unicast
1802 solicitation in the first place. However, including the link-layer
1803 address in this case adds little overhead and eliminates a potential
1804 race condition where the sender deletes the cached link-layer address
1805 prior to receiving a response to a previous solicitation."
1807 ndisc_notify - BOOLEAN
1808 Define mode for notification of address and device changes.
1809 0 - (default): do nothing
1810 1 - Generate unsolicited neighbour advertisements when device is brought
1811 up or hardware address changes.
1813 ndisc_tclass - INTEGER
1814 The IPv6 Traffic Class to use by default when sending IPv6 Neighbor
1815 Discovery (Router Solicitation, Router Advertisement, Neighbor
1816 Solicitation, Neighbor Advertisement, Redirect) messages.
1817 These 8 bits can be interpreted as 6 high order bits holding the DSCP
1818 value and 2 low order bits representing ECN (which you probably want
1822 mldv1_unsolicited_report_interval - INTEGER
1823 The interval in milliseconds in which the next unsolicited
1824 MLDv1 report retransmit will take place.
1825 Default: 10000 (10 seconds)
1827 mldv2_unsolicited_report_interval - INTEGER
1828 The interval in milliseconds in which the next unsolicited
1829 MLDv2 report retransmit will take place.
1830 Default: 1000 (1 second)
1832 force_mld_version - INTEGER
1833 0 - (default) No enforcement of a MLD version, MLDv1 fallback allowed
1834 1 - Enforce to use MLD version 1
1835 2 - Enforce to use MLD version 2
1837 suppress_frag_ndisc - INTEGER
1838 Control RFC 6980 (Security Implications of IPv6 Fragmentation
1839 with IPv6 Neighbor Discovery) behavior:
1840 1 - (default) discard fragmented neighbor discovery packets
1841 0 - allow fragmented neighbor discovery packets
1843 optimistic_dad - BOOLEAN
1844 Whether to perform Optimistic Duplicate Address Detection (RFC 4429).
1845 0: disabled (default)
1848 Optimistic Duplicate Address Detection for the interface will be enabled
1849 if at least one of conf/{all,interface}/optimistic_dad is set to 1,
1850 it will be disabled otherwise.
1852 use_optimistic - BOOLEAN
1853 If enabled, do not classify optimistic addresses as deprecated during
1854 source address selection. Preferred addresses will still be chosen
1855 before optimistic addresses, subject to other ranking in the source
1856 address selection algorithm.
1857 0: disabled (default)
1860 This will be enabled if at least one of
1861 conf/{all,interface}/use_optimistic is set to 1, disabled otherwise.
1863 stable_secret - IPv6 address
1864 This IPv6 address will be used as a secret to generate IPv6
1865 addresses for link-local addresses and autoconfigured
1866 ones. All addresses generated after setting this secret will
1867 be stable privacy ones by default. This can be changed via the
1868 addrgenmode ip-link. conf/default/stable_secret is used as the
1869 secret for the namespace, the interface specific ones can
1870 overwrite that. Writes to conf/all/stable_secret are refused.
1872 It is recommended to generate this secret during installation
1873 of a system and keep it stable after that.
1875 By default the stable secret is unset.
1877 addr_gen_mode - INTEGER
1878 Defines how link-local and autoconf addresses are generated.
1880 0: generate address based on EUI64 (default)
1881 1: do no generate a link-local address, use EUI64 for addresses generated
1883 2: generate stable privacy addresses, using the secret from
1884 stable_secret (RFC7217)
1885 3: generate stable privacy addresses, using a random secret if unset
1887 drop_unicast_in_l2_multicast - BOOLEAN
1888 Drop any unicast IPv6 packets that are received in link-layer
1889 multicast (or broadcast) frames.
1891 By default this is turned off.
1893 drop_unsolicited_na - BOOLEAN
1894 Drop all unsolicited neighbor advertisements, for example if there's
1895 a known good NA proxy on the network and such frames need not be used
1896 (or in the case of 802.11, must not be used to prevent attacks.)
1898 By default this is turned off.
1900 enhanced_dad - BOOLEAN
1901 Include a nonce option in the IPv6 neighbor solicitation messages used for
1902 duplicate address detection per RFC7527. A received DAD NS will only signal
1903 a duplicate address if the nonce is different. This avoids any false
1904 detection of duplicates due to loopback of the NS messages that we send.
1905 The nonce option will be sent on an interface unless both of
1906 conf/{all,interface}/enhanced_dad are set to FALSE.
1911 Limit the maximal rates for sending ICMPv6 packets.
1912 0 to disable any limiting,
1913 otherwise the minimal space between responses in milliseconds.
1916 echo_ignore_all - BOOLEAN
1917 If set non-zero, then the kernel will ignore all ICMP ECHO
1918 requests sent to it over the IPv6 protocol.
1921 xfrm6_gc_thresh - INTEGER
1922 The threshold at which we will start garbage collecting for IPv6
1923 destination cache entries. At twice this value the system will
1924 refuse new allocations.
1928 Pekka Savola <pekkas@netcore.fi>
1929 YOSHIFUJI Hideaki / USAGI Project <yoshfuji@linux-ipv6.org>
1932 /proc/sys/net/bridge/* Variables:
1934 bridge-nf-call-arptables - BOOLEAN
1935 1 : pass bridged ARP traffic to arptables' FORWARD chain.
1939 bridge-nf-call-iptables - BOOLEAN
1940 1 : pass bridged IPv4 traffic to iptables' chains.
1944 bridge-nf-call-ip6tables - BOOLEAN
1945 1 : pass bridged IPv6 traffic to ip6tables' chains.
1949 bridge-nf-filter-vlan-tagged - BOOLEAN
1950 1 : pass bridged vlan-tagged ARP/IP/IPv6 traffic to {arp,ip,ip6}tables.
1954 bridge-nf-filter-pppoe-tagged - BOOLEAN
1955 1 : pass bridged pppoe-tagged IP/IPv6 traffic to {ip,ip6}tables.
1959 bridge-nf-pass-vlan-input-dev - BOOLEAN
1960 1: if bridge-nf-filter-vlan-tagged is enabled, try to find a vlan
1961 interface on the bridge and set the netfilter input device to the vlan.
1962 This allows use of e.g. "iptables -i br0.1" and makes the REDIRECT
1963 target work with vlan-on-top-of-bridge interfaces. When no matching
1964 vlan interface is found, or this switch is off, the input device is
1965 set to the bridge interface.
1966 0: disable bridge netfilter vlan interface lookup.
1969 proc/sys/net/sctp/* Variables:
1971 addip_enable - BOOLEAN
1972 Enable or disable extension of Dynamic Address Reconfiguration
1973 (ADD-IP) functionality specified in RFC5061. This extension provides
1974 the ability to dynamically add and remove new addresses for the SCTP
1977 1: Enable extension.
1979 0: Disable extension.
1984 Enable or disable pf (pf is short for potentially failed) state. A value
1985 of pf_retrans > path_max_retrans also disables pf state. That is, one of
1986 both pf_enable and pf_retrans > path_max_retrans can disable pf state.
1987 Since pf_retrans and path_max_retrans can be changed by userspace
1988 application, sometimes user expects to disable pf state by the value of
1989 pf_retrans > path_max_retrans, but occasionally the value of pf_retrans
1990 or path_max_retrans is changed by the user application, this pf state is
1991 enabled. As such, it is necessary to add this to dynamically enable
1992 and disable pf state. See:
1993 https://datatracker.ietf.org/doc/draft-ietf-tsvwg-sctp-failover for
2002 addip_noauth_enable - BOOLEAN
2003 Dynamic Address Reconfiguration (ADD-IP) requires the use of
2004 authentication to protect the operations of adding or removing new
2005 addresses. This requirement is mandated so that unauthorized hosts
2006 would not be able to hijack associations. However, older
2007 implementations may not have implemented this requirement while
2008 allowing the ADD-IP extension. For reasons of interoperability,
2009 we provide this variable to control the enforcement of the
2010 authentication requirement.
2012 1: Allow ADD-IP extension to be used without authentication. This
2013 should only be set in a closed environment for interoperability
2014 with older implementations.
2016 0: Enforce the authentication requirement
2020 auth_enable - BOOLEAN
2021 Enable or disable Authenticated Chunks extension. This extension
2022 provides the ability to send and receive authenticated chunks and is
2023 required for secure operation of Dynamic Address Reconfiguration
2026 1: Enable this extension.
2027 0: Disable this extension.
2031 prsctp_enable - BOOLEAN
2032 Enable or disable the Partial Reliability extension (RFC3758) which
2033 is used to notify peers that a given DATA should no longer be expected.
2041 The limit of the number of new packets that can be initially sent. It
2042 controls how bursty the generated traffic can be.
2046 association_max_retrans - INTEGER
2047 Set the maximum number for retransmissions that an association can
2048 attempt deciding that the remote end is unreachable. If this value
2049 is exceeded, the association is terminated.
2053 max_init_retransmits - INTEGER
2054 The maximum number of retransmissions of INIT and COOKIE-ECHO chunks
2055 that an association will attempt before declaring the destination
2056 unreachable and terminating.
2060 path_max_retrans - INTEGER
2061 The maximum number of retransmissions that will be attempted on a given
2062 path. Once this threshold is exceeded, the path is considered
2063 unreachable, and new traffic will use a different path when the
2064 association is multihomed.
2068 pf_retrans - INTEGER
2069 The number of retransmissions that will be attempted on a given path
2070 before traffic is redirected to an alternate transport (should one
2071 exist). Note this is distinct from path_max_retrans, as a path that
2072 passes the pf_retrans threshold can still be used. Its only
2073 deprioritized when a transmission path is selected by the stack. This
2074 setting is primarily used to enable fast failover mechanisms without
2075 having to reduce path_max_retrans to a very low value. See:
2076 http://www.ietf.org/id/draft-nishida-tsvwg-sctp-failover-05.txt
2077 for details. Note also that a value of pf_retrans > path_max_retrans
2078 disables this feature. Since both pf_retrans and path_max_retrans can
2079 be changed by userspace application, a variable pf_enable is used to
2084 rto_initial - INTEGER
2085 The initial round trip timeout value in milliseconds that will be used
2086 in calculating round trip times. This is the initial time interval
2087 for retransmissions.
2092 The maximum value (in milliseconds) of the round trip timeout. This
2093 is the largest time interval that can elapse between retransmissions.
2098 The minimum value (in milliseconds) of the round trip timeout. This
2099 is the smallest time interval the can elapse between retransmissions.
2103 hb_interval - INTEGER
2104 The interval (in milliseconds) between HEARTBEAT chunks. These chunks
2105 are sent at the specified interval on idle paths to probe the state of
2106 a given path between 2 associations.
2110 sack_timeout - INTEGER
2111 The amount of time (in milliseconds) that the implementation will wait
2116 valid_cookie_life - INTEGER
2117 The default lifetime of the SCTP cookie (in milliseconds). The cookie
2118 is used during association establishment.
2122 cookie_preserve_enable - BOOLEAN
2123 Enable or disable the ability to extend the lifetime of the SCTP cookie
2124 that is used during the establishment phase of SCTP association
2126 1: Enable cookie lifetime extension.
2131 cookie_hmac_alg - STRING
2132 Select the hmac algorithm used when generating the cookie value sent by
2133 a listening sctp socket to a connecting client in the INIT-ACK chunk.
2138 Ability to assign md5 or sha1 as the selected alg is predicated on the
2139 configuration of those algorithms at build time (CONFIG_CRYPTO_MD5 and
2140 CONFIG_CRYPTO_SHA1).
2142 Default: Dependent on configuration. MD5 if available, else SHA1 if
2143 available, else none.
2145 rcvbuf_policy - INTEGER
2146 Determines if the receive buffer is attributed to the socket or to
2147 association. SCTP supports the capability to create multiple
2148 associations on a single socket. When using this capability, it is
2149 possible that a single stalled association that's buffering a lot
2150 of data may block other associations from delivering their data by
2151 consuming all of the receive buffer space. To work around this,
2152 the rcvbuf_policy could be set to attribute the receiver buffer space
2153 to each association instead of the socket. This prevents the described
2156 1: rcvbuf space is per association
2157 0: rcvbuf space is per socket
2161 sndbuf_policy - INTEGER
2162 Similar to rcvbuf_policy above, this applies to send buffer space.
2164 1: Send buffer is tracked per association
2165 0: Send buffer is tracked per socket.
2169 sctp_mem - vector of 3 INTEGERs: min, pressure, max
2170 Number of pages allowed for queueing by all SCTP sockets.
2172 min: Below this number of pages SCTP is not bothered about its
2173 memory appetite. When amount of memory allocated by SCTP exceeds
2174 this number, SCTP starts to moderate memory usage.
2176 pressure: This value was introduced to follow format of tcp_mem.
2178 max: Number of pages allowed for queueing by all SCTP sockets.
2180 Default is calculated at boot time from amount of available memory.
2182 sctp_rmem - vector of 3 INTEGERs: min, default, max
2183 Only the first value ("min") is used, "default" and "max" are
2186 min: Minimal size of receive buffer used by SCTP socket.
2187 It is guaranteed to each SCTP socket (but not association) even
2188 under moderate memory pressure.
2192 sctp_wmem - vector of 3 INTEGERs: min, default, max
2193 Currently this tunable has no effect.
2195 addr_scope_policy - INTEGER
2196 Control IPv4 address scoping - draft-stewart-tsvwg-sctp-ipv4-00
2198 0 - Disable IPv4 address scoping
2199 1 - Enable IPv4 address scoping
2200 2 - Follow draft but allow IPv4 private addresses
2201 3 - Follow draft but allow IPv4 link local addresses
2206 /proc/sys/net/core/*
2207 Please see: Documentation/sysctl/net.txt for descriptions of these entries.
2210 /proc/sys/net/unix/*
2211 max_dgram_qlen - INTEGER
2212 The maximum length of dgram socket receive queue