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
16 ip_no_pmtu_disc - BOOLEAN
17 Disable Path MTU Discovery.
21 default 562 - minimum discovered Path MTU
24 Time, in seconds, that cached PMTU information is kept.
27 The advertised MSS depends on the first hop route MTU, but will
28 never be lower than this setting.
32 ipfrag_high_thresh - INTEGER
33 Maximum memory used to reassemble IP fragments. When
34 ipfrag_high_thresh bytes of memory is allocated for this purpose,
35 the fragment handler will toss packets until ipfrag_low_thresh
38 ipfrag_low_thresh - INTEGER
39 See ipfrag_high_thresh
42 Time in seconds to keep an IP fragment in memory.
44 ipfrag_secret_interval - INTEGER
45 Regeneration interval (in seconds) of the hash secret (or lifetime
46 for the hash secret) for IP fragments.
49 ipfrag_max_dist - INTEGER
50 ipfrag_max_dist is a non-negative integer value which defines the
51 maximum "disorder" which is allowed among fragments which share a
52 common IP source address. Note that reordering of packets is
53 not unusual, but if a large number of fragments arrive from a source
54 IP address while a particular fragment queue remains incomplete, it
55 probably indicates that one or more fragments belonging to that queue
56 have been lost. When ipfrag_max_dist is positive, an additional check
57 is done on fragments before they are added to a reassembly queue - if
58 ipfrag_max_dist (or more) fragments have arrived from a particular IP
59 address between additions to any IP fragment queue using that source
60 address, it's presumed that one or more fragments in the queue are
61 lost. The existing fragment queue will be dropped, and a new one
62 started. An ipfrag_max_dist value of zero disables this check.
64 Using a very small value, e.g. 1 or 2, for ipfrag_max_dist can
65 result in unnecessarily dropping fragment queues when normal
66 reordering of packets occurs, which could lead to poor application
67 performance. Using a very large value, e.g. 50000, increases the
68 likelihood of incorrectly reassembling IP fragments that originate
69 from different IP datagrams, which could result in data corruption.
74 inet_peer_threshold - INTEGER
75 The approximate size of the storage. Starting from this threshold
76 entries will be thrown aggressively. This threshold also determines
77 entries' time-to-live and time intervals between garbage collection
78 passes. More entries, less time-to-live, less GC interval.
80 inet_peer_minttl - INTEGER
81 Minimum time-to-live of entries. Should be enough to cover fragment
82 time-to-live on the reassembling side. This minimum time-to-live is
83 guaranteed if the pool size is less than inet_peer_threshold.
84 Measured in jiffies(1).
86 inet_peer_maxttl - INTEGER
87 Maximum time-to-live of entries. Unused entries will expire after
88 this period of time if there is no memory pressure on the pool (i.e.
89 when the number of entries in the pool is very small).
90 Measured in jiffies(1).
92 inet_peer_gc_mintime - INTEGER
93 Minimum interval between garbage collection passes. This interval is
94 in effect under high memory pressure on the pool.
95 Measured in jiffies(1).
97 inet_peer_gc_maxtime - INTEGER
98 Minimum interval between garbage collection passes. This interval is
99 in effect under low (or absent) memory pressure on the pool.
100 Measured in jiffies(1).
105 Limit of socket listen() backlog, known in userspace as SOMAXCONN.
106 Defaults to 128. See also tcp_max_syn_backlog for additional tuning
110 Controls Appropriate Byte Count (ABC) defined in RFC3465.
111 ABC is a way of increasing congestion window (cwnd) more slowly
112 in response to partial acknowledgments.
114 0 increase cwnd once per acknowledgment (no ABC)
115 1 increase cwnd once per acknowledgment of full sized segment
116 2 allow increase cwnd by two if acknowledgment is
117 of two segments to compensate for delayed acknowledgments.
120 tcp_abort_on_overflow - BOOLEAN
121 If listening service is too slow to accept new connections,
122 reset them. Default state is FALSE. It means that if overflow
123 occurred due to a burst, connection will recover. Enable this
124 option _only_ if you are really sure that listening daemon
125 cannot be tuned to accept connections faster. Enabling this
126 option can harm clients of your server.
128 tcp_adv_win_scale - INTEGER
129 Count buffering overhead as bytes/2^tcp_adv_win_scale
130 (if tcp_adv_win_scale > 0) or bytes-bytes/2^(-tcp_adv_win_scale),
134 tcp_allowed_congestion_control - STRING
135 Show/set the congestion control choices available to non-privileged
136 processes. The list is a subset of those listed in
137 tcp_available_congestion_control.
138 Default is "reno" and the default setting (tcp_congestion_control).
140 tcp_app_win - INTEGER
141 Reserve max(window/2^tcp_app_win, mss) of window for application
142 buffer. Value 0 is special, it means that nothing is reserved.
145 tcp_available_congestion_control - STRING
146 Shows the available congestion control choices that are registered.
147 More congestion control algorithms may be available as modules,
150 tcp_base_mss - INTEGER
151 The initial value of search_low to be used by Packetization Layer
152 Path MTU Discovery (MTU probing). If MTU probing is enabled,
153 this is the inital MSS used by the connection.
155 tcp_congestion_control - STRING
156 Set the congestion control algorithm to be used for new
157 connections. The algorithm "reno" is always available, but
158 additional choices may be available based on kernel configuration.
159 Default is set as part of kernel configuration.
162 Allows TCP to send "duplicate" SACKs.
165 Enable Explicit Congestion Notification in TCP.
168 Enable FACK congestion avoidance and fast retransmission.
169 The value is not used, if tcp_sack is not enabled.
171 tcp_fin_timeout - INTEGER
172 Time to hold socket in state FIN-WAIT-2, if it was closed
173 by our side. Peer can be broken and never close its side,
174 or even died unexpectedly. Default value is 60sec.
175 Usual value used in 2.2 was 180 seconds, you may restore
176 it, but remember that if your machine is even underloaded WEB server,
177 you risk to overflow memory with kilotons of dead sockets,
178 FIN-WAIT-2 sockets are less dangerous than FIN-WAIT-1,
179 because they eat maximum 1.5K of memory, but they tend
180 to live longer. Cf. tcp_max_orphans.
183 Enables Forward RTO-Recovery (F-RTO) defined in RFC4138.
184 F-RTO is an enhanced recovery algorithm for TCP retransmission
185 timeouts. It is particularly beneficial in wireless environments
186 where packet loss is typically due to random radio interference
187 rather than intermediate router congestion. F-RTO is sender-side
188 only modification. Therefore it does not require any support from
189 the peer, but in a typical case, however, where wireless link is
190 the local access link and most of the data flows downlink, the
191 faraway servers should have F-RTO enabled to take advantage of it.
192 If set to 1, basic version is enabled. 2 enables SACK enhanced
193 F-RTO if flow uses SACK. The basic version can be used also when
194 SACK is in use though scenario(s) with it exists where F-RTO
195 interacts badly with the packet counting of the SACK enabled TCP
198 tcp_frto_response - INTEGER
199 When F-RTO has detected that a TCP retransmission timeout was
200 spurious (i.e, the timeout would have been avoided had TCP set a
201 longer retransmission timeout), TCP has several options what to do
202 next. Possible values are:
203 0 Rate halving based; a smooth and conservative response,
204 results in halved cwnd and ssthresh after one RTT
205 1 Very conservative response; not recommended because even
206 though being valid, it interacts poorly with the rest of
207 Linux TCP, halves cwnd and ssthresh immediately
208 2 Aggressive response; undoes congestion control measures
209 that are now known to be unnecessary (ignoring the
210 possibility of a lost retransmission that would require
211 TCP to be more cautious), cwnd and ssthresh are restored
212 to the values prior timeout
213 Default: 0 (rate halving based)
215 tcp_keepalive_time - INTEGER
216 How often TCP sends out keepalive messages when keepalive is enabled.
219 tcp_keepalive_probes - INTEGER
220 How many keepalive probes TCP sends out, until it decides that the
221 connection is broken. Default value: 9.
223 tcp_keepalive_intvl - INTEGER
224 How frequently the probes are send out. Multiplied by
225 tcp_keepalive_probes it is time to kill not responding connection,
226 after probes started. Default value: 75sec i.e. connection
227 will be aborted after ~11 minutes of retries.
229 tcp_low_latency - BOOLEAN
230 If set, the TCP stack makes decisions that prefer lower
231 latency as opposed to higher throughput. By default, this
232 option is not set meaning that higher throughput is preferred.
233 An example of an application where this default should be
234 changed would be a Beowulf compute cluster.
237 tcp_max_orphans - INTEGER
238 Maximal number of TCP sockets not attached to any user file handle,
239 held by system. If this number is exceeded orphaned connections are
240 reset immediately and warning is printed. This limit exists
241 only to prevent simple DoS attacks, you _must_ not rely on this
242 or lower the limit artificially, but rather increase it
243 (probably, after increasing installed memory),
244 if network conditions require more than default value,
245 and tune network services to linger and kill such states
246 more aggressively. Let me to remind again: each orphan eats
247 up to ~64K of unswappable memory.
249 tcp_max_syn_backlog - INTEGER
250 Maximal number of remembered connection requests, which are
251 still did not receive an acknowledgment from connecting client.
252 Default value is 1024 for systems with more than 128Mb of memory,
253 and 128 for low memory machines. If server suffers of overload,
254 try to increase this number.
256 tcp_max_tw_buckets - INTEGER
257 Maximal number of timewait sockets held by system simultaneously.
258 If this number is exceeded time-wait socket is immediately destroyed
259 and warning is printed. This limit exists only to prevent
260 simple DoS attacks, you _must_ not lower the limit artificially,
261 but rather increase it (probably, after increasing installed memory),
262 if network conditions require more than default value.
264 tcp_mem - vector of 3 INTEGERs: min, pressure, max
265 min: below this number of pages TCP is not bothered about its
268 pressure: when amount of memory allocated by TCP exceeds this number
269 of pages, TCP moderates its memory consumption and enters memory
270 pressure mode, which is exited when memory consumption falls
273 max: number of pages allowed for queueing by all TCP sockets.
275 Defaults are calculated at boot time from amount of available
278 tcp_moderate_rcvbuf - BOOLEAN
279 If set, TCP performs receive buffer autotuning, attempting to
280 automatically size the buffer (no greater than tcp_rmem[2]) to
281 match the size required by the path for full throughput. Enabled by
284 tcp_mtu_probing - INTEGER
285 Controls TCP Packetization-Layer Path MTU Discovery. Takes three
288 1 - Disabled by default, enabled when an ICMP black hole detected
289 2 - Always enabled, use initial MSS of tcp_base_mss.
291 tcp_no_metrics_save - BOOLEAN
292 By default, TCP saves various connection metrics in the route cache
293 when the connection closes, so that connections established in the
294 near future can use these to set initial conditions. Usually, this
295 increases overall performance, but may sometimes cause performance
296 degradation. If set, TCP will not cache metrics on closing
299 tcp_orphan_retries - INTEGER
300 How may times to retry before killing TCP connection, closed
301 by our side. Default value 7 corresponds to ~50sec-16min
302 depending on RTO. If you machine is loaded WEB server,
303 you should think about lowering this value, such sockets
304 may consume significant resources. Cf. tcp_max_orphans.
306 tcp_reordering - INTEGER
307 Maximal reordering of packets in a TCP stream.
310 tcp_retrans_collapse - BOOLEAN
311 Bug-to-bug compatibility with some broken printers.
312 On retransmit try to send bigger packets to work around bugs in
315 tcp_retries1 - INTEGER
316 How many times to retry before deciding that something is wrong
317 and it is necessary to report this suspicion to network layer.
318 Minimal RFC value is 3, it is default, which corresponds
319 to ~3sec-8min depending on RTO.
321 tcp_retries2 - INTEGER
322 How may times to retry before killing alive TCP connection.
323 RFC1122 says that the limit should be longer than 100 sec.
324 It is too small number. Default value 15 corresponds to ~13-30min
327 tcp_rfc1337 - BOOLEAN
328 If set, the TCP stack behaves conforming to RFC1337. If unset,
329 we are not conforming to RFC, but prevent TCP TIME_WAIT
333 tcp_rmem - vector of 3 INTEGERs: min, default, max
334 min: Minimal size of receive buffer used by TCP sockets.
335 It is guaranteed to each TCP socket, even under moderate memory
339 default: default size of receive buffer used by TCP sockets.
340 This value overrides net.core.rmem_default used by other protocols.
341 Default: 87380 bytes. This value results in window of 65535 with
342 default setting of tcp_adv_win_scale and tcp_app_win:0 and a bit
343 less for default tcp_app_win. See below about these variables.
345 max: maximal size of receive buffer allowed for automatically
346 selected receiver buffers for TCP socket. This value does not override
347 net.core.rmem_max, "static" selection via SO_RCVBUF does not use this.
348 Default: 87380*2 bytes.
351 Enable select acknowledgments (SACKS).
353 tcp_slow_start_after_idle - BOOLEAN
354 If set, provide RFC2861 behavior and time out the congestion
355 window after an idle period. An idle period is defined at
356 the current RTO. If unset, the congestion window will not
357 be timed out after an idle period.
361 Use the Host requirements interpretation of the TCP urg pointer field.
362 Most hosts use the older BSD interpretation, so if you turn this on
363 Linux might not communicate correctly with them.
366 tcp_synack_retries - INTEGER
367 Number of times SYNACKs for a passive TCP connection attempt will
368 be retransmitted. Should not be higher than 255. Default value
369 is 5, which corresponds to ~180seconds.
371 tcp_syncookies - BOOLEAN
372 Only valid when the kernel was compiled with CONFIG_SYNCOOKIES
373 Send out syncookies when the syn backlog queue of a socket
374 overflows. This is to prevent against the common 'syn flood attack'
377 Note, that syncookies is fallback facility.
378 It MUST NOT be used to help highly loaded servers to stand
379 against legal connection rate. If you see synflood warnings
380 in your logs, but investigation shows that they occur
381 because of overload with legal connections, you should tune
382 another parameters until this warning disappear.
383 See: tcp_max_syn_backlog, tcp_synack_retries, tcp_abort_on_overflow.
385 syncookies seriously violate TCP protocol, do not allow
386 to use TCP extensions, can result in serious degradation
387 of some services (f.e. SMTP relaying), visible not by you,
388 but your clients and relays, contacting you. While you see
389 synflood warnings in logs not being really flooded, your server
390 is seriously misconfigured.
392 tcp_syn_retries - INTEGER
393 Number of times initial SYNs for an active TCP connection attempt
394 will be retransmitted. Should not be higher than 255. Default value
395 is 5, which corresponds to ~180seconds.
397 tcp_timestamps - BOOLEAN
398 Enable timestamps as defined in RFC1323.
400 tcp_tso_win_divisor - INTEGER
401 This allows control over what percentage of the congestion window
402 can be consumed by a single TSO frame.
403 The setting of this parameter is a choice between burstiness and
404 building larger TSO frames.
407 tcp_tw_recycle - BOOLEAN
408 Enable fast recycling TIME-WAIT sockets. Default value is 0.
409 It should not be changed without advice/request of technical
412 tcp_tw_reuse - BOOLEAN
413 Allow to reuse TIME-WAIT sockets for new connections when it is
414 safe from protocol viewpoint. Default value is 0.
415 It should not be changed without advice/request of technical
418 tcp_window_scaling - BOOLEAN
419 Enable window scaling as defined in RFC1323.
421 tcp_wmem - vector of 3 INTEGERs: min, default, max
422 min: Amount of memory reserved for send buffers for TCP socket.
423 Each TCP socket has rights to use it due to fact of its birth.
426 default: Amount of memory allowed for send buffers for TCP socket
427 by default. This value overrides net.core.wmem_default used
428 by other protocols, it is usually lower than net.core.wmem_default.
431 max: Maximal amount of memory allowed for automatically selected
432 send buffers for TCP socket. This value does not override
433 net.core.wmem_max, "static" selection via SO_SNDBUF does not use this.
436 tcp_workaround_signed_windows - BOOLEAN
437 If set, assume no receipt of a window scaling option means the
438 remote TCP is broken and treats the window as a signed quantity.
439 If unset, assume the remote TCP is not broken even if we do
440 not receive a window scaling option from them.
443 tcp_dma_copybreak - INTEGER
444 Lower limit, in bytes, of the size of socket reads that will be
445 offloaded to a DMA copy engine, if one is present in the system
446 and CONFIG_NET_DMA is enabled.
451 udp_mem - vector of 3 INTEGERs: min, pressure, max
452 Number of pages allowed for queueing by all UDP sockets.
454 min: Below this number of pages UDP is not bothered about its
455 memory appetite. When amount of memory allocated by UDP exceeds
456 this number, UDP starts to moderate memory usage.
458 pressure: This value was introduced to follow format of tcp_mem.
460 max: Number of pages allowed for queueing by all UDP sockets.
462 Default is calculated at boot time from amount of available memory.
464 udp_rmem_min - INTEGER
465 Minimal size of receive buffer used by UDP sockets in moderation.
466 Each UDP socket is able to use the size for receiving data, even if
467 total pages of UDP sockets exceed udp_mem pressure. The unit is byte.
470 udp_wmem_min - INTEGER
471 Minimal size of send buffer used by UDP sockets in moderation.
472 Each UDP socket is able to use the size for sending data, even if
473 total pages of UDP sockets exceed udp_mem pressure. The unit is byte.
478 cipso_cache_enable - BOOLEAN
479 If set, enable additions to and lookups from the CIPSO label mapping
480 cache. If unset, additions are ignored and lookups always result in a
481 miss. However, regardless of the setting the cache is still
482 invalidated when required when means you can safely toggle this on and
483 off and the cache will always be "safe".
486 cipso_cache_bucket_size - INTEGER
487 The CIPSO label cache consists of a fixed size hash table with each
488 hash bucket containing a number of cache entries. This variable limits
489 the number of entries in each hash bucket; the larger the value the
490 more CIPSO label mappings that can be cached. When the number of
491 entries in a given hash bucket reaches this limit adding new entries
492 causes the oldest entry in the bucket to be removed to make room.
495 cipso_rbm_optfmt - BOOLEAN
496 Enable the "Optimized Tag 1 Format" as defined in section 3.4.2.6 of
497 the CIPSO draft specification (see Documentation/netlabel for details).
498 This means that when set the CIPSO tag will be padded with empty
499 categories in order to make the packet data 32-bit aligned.
502 cipso_rbm_structvalid - BOOLEAN
503 If set, do a very strict check of the CIPSO option when
504 ip_options_compile() is called. If unset, relax the checks done during
505 ip_options_compile(). Either way is "safe" as errors are caught else
506 where in the CIPSO processing code but setting this to 0 (False) should
507 result in less work (i.e. it should be faster) but could cause problems
508 with other implementations that require strict checking.
513 ip_local_port_range - 2 INTEGERS
514 Defines the local port range that is used by TCP and UDP to
515 choose the local port. The first number is the first, the
516 second the last local port number. Default value depends on
517 amount of memory available on the system:
519 < 128Mb 1024-4999 or even less.
520 This number defines number of active connections, which this
521 system can issue simultaneously to systems not supporting
522 TCP extensions (timestamps). With tcp_tw_recycle enabled
523 (i.e. by default) range 1024-4999 is enough to issue up to
524 2000 connections per second to systems supporting timestamps.
526 ip_nonlocal_bind - BOOLEAN
527 If set, allows processes to bind() to non-local IP addresses,
528 which can be quite useful - but may break some applications.
532 If set non-zero, enables support for dynamic addresses.
533 If set to a non-zero value larger than 1, a kernel log
534 message will be printed when dynamic address rewriting
538 icmp_echo_ignore_all - BOOLEAN
539 If set non-zero, then the kernel will ignore all ICMP ECHO
543 icmp_echo_ignore_broadcasts - BOOLEAN
544 If set non-zero, then the kernel will ignore all ICMP ECHO and
545 TIMESTAMP requests sent to it via broadcast/multicast.
548 icmp_ratelimit - INTEGER
549 Limit the maximal rates for sending ICMP packets whose type matches
550 icmp_ratemask (see below) to specific targets.
551 0 to disable any limiting,
552 otherwise the minimal space between responses in milliseconds.
555 icmp_ratemask - INTEGER
556 Mask made of ICMP types for which rates are being limited.
557 Significant bits: IHGFEDCBA9876543210
558 Default mask: 0000001100000011000 (6168)
560 Bit definitions (see include/linux/icmp.h):
562 3 Destination Unreachable *
567 C Parameter Problem *
572 H Address Mask Request
575 * These are rate limited by default (see default mask above)
577 icmp_ignore_bogus_error_responses - BOOLEAN
578 Some routers violate RFC1122 by sending bogus responses to broadcast
579 frames. Such violations are normally logged via a kernel warning.
580 If this is set to TRUE, the kernel will not give such warnings, which
581 will avoid log file clutter.
584 icmp_errors_use_inbound_ifaddr - BOOLEAN
586 If zero, icmp error messages are sent with the primary address of
587 the exiting interface.
589 If non-zero, the message will be sent with the primary address of
590 the interface that received the packet that caused the icmp error.
591 This is the behaviour network many administrators will expect from
592 a router. And it can make debugging complicated network layouts
595 Note that if no primary address exists for the interface selected,
596 then the primary address of the first non-loopback interface that
597 has one will be used regardless of this setting.
601 igmp_max_memberships - INTEGER
602 Change the maximum number of multicast groups we can subscribe to.
605 conf/interface/* changes special settings per interface (where "interface" is
606 the name of your network interface)
607 conf/all/* is special, changes the settings for all interfaces
610 log_martians - BOOLEAN
611 Log packets with impossible addresses to kernel log.
612 log_martians for the interface will be enabled if at least one of
613 conf/{all,interface}/log_martians is set to TRUE,
614 it will be disabled otherwise
616 accept_redirects - BOOLEAN
617 Accept ICMP redirect messages.
618 accept_redirects for the interface will be enabled if:
619 - both conf/{all,interface}/accept_redirects are TRUE in the case forwarding
620 for the interface is enabled
622 - at least one of conf/{all,interface}/accept_redirects is TRUE in the case
623 forwarding for the interface is disabled
624 accept_redirects for the interface will be disabled otherwise
629 Enable IP forwarding on this interface.
631 mc_forwarding - BOOLEAN
632 Do multicast routing. The kernel needs to be compiled with CONFIG_MROUTE
633 and a multicast routing daemon is required.
634 conf/all/mc_forwarding must also be set to TRUE to enable multicast routing
638 Integer value used to differentiate the devices by the medium they
639 are attached to. Two devices can have different id values when
640 the broadcast packets are received only on one of them.
641 The default value 0 means that the device is the only interface
642 to its medium, value of -1 means that medium is not known.
644 Currently, it is used to change the proxy_arp behavior:
645 the proxy_arp feature is enabled for packets forwarded between
646 two devices attached to different media.
650 proxy_arp for the interface will be enabled if at least one of
651 conf/{all,interface}/proxy_arp is set to TRUE,
652 it will be disabled otherwise
654 shared_media - BOOLEAN
655 Send(router) or accept(host) RFC1620 shared media redirects.
656 Overrides ip_secure_redirects.
657 shared_media for the interface will be enabled if at least one of
658 conf/{all,interface}/shared_media is set to TRUE,
659 it will be disabled otherwise
662 secure_redirects - BOOLEAN
663 Accept ICMP redirect messages only for gateways,
664 listed in default gateway list.
665 secure_redirects for the interface will be enabled if at least one of
666 conf/{all,interface}/secure_redirects is set to TRUE,
667 it will be disabled otherwise
670 send_redirects - BOOLEAN
671 Send redirects, if router.
672 send_redirects for the interface will be enabled if at least one of
673 conf/{all,interface}/send_redirects is set to TRUE,
674 it will be disabled otherwise
677 bootp_relay - BOOLEAN
678 Accept packets with source address 0.b.c.d destined
679 not to this host as local ones. It is supposed, that
680 BOOTP relay daemon will catch and forward such packets.
681 conf/all/bootp_relay must also be set to TRUE to enable BOOTP relay
686 accept_source_route - BOOLEAN
687 Accept packets with SRR option.
688 conf/all/accept_source_route must also be set to TRUE to accept packets
689 with SRR option on the interface
690 default TRUE (router)
694 1 - do source validation by reversed path, as specified in RFC1812
695 Recommended option for single homed hosts and stub network
696 routers. Could cause troubles for complicated (not loop free)
697 networks running a slow unreliable protocol (sort of RIP),
698 or using static routes.
700 0 - No source validation.
702 conf/all/rp_filter must also be set to TRUE to do source validation
705 Default value is 0. Note that some distributions enable it
709 1 - Allows you to have multiple network interfaces on the same
710 subnet, and have the ARPs for each interface be answered
711 based on whether or not the kernel would route a packet from
712 the ARP'd IP out that interface (therefore you must use source
713 based routing for this to work). In other words it allows control
714 of which cards (usually 1) will respond to an arp request.
716 0 - (default) The kernel can respond to arp requests with addresses
717 from other interfaces. This may seem wrong but it usually makes
718 sense, because it increases the chance of successful communication.
719 IP addresses are owned by the complete host on Linux, not by
720 particular interfaces. Only for more complex setups like load-
721 balancing, does this behaviour cause problems.
723 arp_filter for the interface will be enabled if at least one of
724 conf/{all,interface}/arp_filter is set to TRUE,
725 it will be disabled otherwise
727 arp_announce - INTEGER
728 Define different restriction levels for announcing the local
729 source IP address from IP packets in ARP requests sent on
731 0 - (default) Use any local address, configured on any interface
732 1 - Try to avoid local addresses that are not in the target's
733 subnet for this interface. This mode is useful when target
734 hosts reachable via this interface require the source IP
735 address in ARP requests to be part of their logical network
736 configured on the receiving interface. When we generate the
737 request we will check all our subnets that include the
738 target IP and will preserve the source address if it is from
739 such subnet. If there is no such subnet we select source
740 address according to the rules for level 2.
741 2 - Always use the best local address for this target.
742 In this mode we ignore the source address in the IP packet
743 and try to select local address that we prefer for talks with
744 the target host. Such local address is selected by looking
745 for primary IP addresses on all our subnets on the outgoing
746 interface that include the target IP address. If no suitable
747 local address is found we select the first local address
748 we have on the outgoing interface or on all other interfaces,
749 with the hope we will receive reply for our request and
750 even sometimes no matter the source IP address we announce.
752 The max value from conf/{all,interface}/arp_announce is used.
754 Increasing the restriction level gives more chance for
755 receiving answer from the resolved target while decreasing
756 the level announces more valid sender's information.
759 Define different modes for sending replies in response to
760 received ARP requests that resolve local target IP addresses:
761 0 - (default): reply for any local target IP address, configured
763 1 - reply only if the target IP address is local address
764 configured on the incoming interface
765 2 - reply only if the target IP address is local address
766 configured on the incoming interface and both with the
767 sender's IP address are part from same subnet on this interface
768 3 - do not reply for local addresses configured with scope host,
769 only resolutions for global and link addresses are replied
771 8 - do not reply for all local addresses
773 The max value from conf/{all,interface}/arp_ignore is used
774 when ARP request is received on the {interface}
777 Define behavior when gratuitous arp replies are received:
778 0 - drop gratuitous arp frames
779 1 - accept gratuitous arp frames
781 app_solicit - INTEGER
782 The maximum number of probes to send to the user space ARP daemon
783 via netlink before dropping back to multicast probes (see
784 mcast_solicit). Defaults to 0.
786 disable_policy - BOOLEAN
787 Disable IPSEC policy (SPD) for this interface
789 disable_xfrm - BOOLEAN
790 Disable IPSEC encryption on this interface, whatever the policy
795 Allows you to write a number, which can be used as required.
798 (1) Jiffie: internal timeunit for the kernel. On the i386 1/100s, on the
799 Alpha 1/1024s. See the HZ define in /usr/include/asm/param.h for the exact
800 value on your system.
809 delon.nicolas@wanadoo.fr
814 /proc/sys/net/ipv6/* Variables:
816 IPv6 has no global variables such as tcp_*. tcp_* settings under ipv4/ also
817 apply to IPv6 [XXX?].
820 Default value for IPV6_V6ONLY socket option,
821 which restricts use of the IPv6 socket to IPv6 communication
823 TRUE: disable IPv4-mapped address feature
824 FALSE: enable IPv4-mapped address feature
826 Default: FALSE (as specified in RFC2553bis)
830 ip6frag_high_thresh - INTEGER
831 Maximum memory used to reassemble IPv6 fragments. When
832 ip6frag_high_thresh bytes of memory is allocated for this purpose,
833 the fragment handler will toss packets until ip6frag_low_thresh
836 ip6frag_low_thresh - INTEGER
837 See ip6frag_high_thresh
839 ip6frag_time - INTEGER
840 Time in seconds to keep an IPv6 fragment in memory.
842 ip6frag_secret_interval - INTEGER
843 Regeneration interval (in seconds) of the hash secret (or lifetime
844 for the hash secret) for IPv6 fragments.
848 Change the interface-specific default settings.
852 Change all the interface-specific settings.
854 [XXX: Other special features than forwarding?]
856 conf/all/forwarding - BOOLEAN
857 Enable global IPv6 forwarding between all interfaces.
859 IPv4 and IPv6 work differently here; e.g. netfilter must be used
860 to control which interfaces may forward packets and which not.
862 This also sets all interfaces' Host/Router setting
863 'forwarding' to the specified value. See below for details.
865 This referred to as global forwarding.
871 Change special settings per interface.
873 The functional behaviour for certain settings is different
874 depending on whether local forwarding is enabled or not.
877 Accept Router Advertisements; autoconfigure using them.
879 Functional default: enabled if local forwarding is disabled.
880 disabled if local forwarding is enabled.
882 accept_ra_defrtr - BOOLEAN
883 Learn default router in Router Advertisement.
885 Functional default: enabled if accept_ra is enabled.
886 disabled if accept_ra is disabled.
888 accept_ra_pinfo - BOOLEAN
889 Learn Prefix Information in Router Advertisement.
891 Functional default: enabled if accept_ra is enabled.
892 disabled if accept_ra is disabled.
894 accept_ra_rt_info_max_plen - INTEGER
895 Maximum prefix length of Route Information in RA.
897 Route Information w/ prefix larger than or equal to this
898 variable shall be ignored.
900 Functional default: 0 if accept_ra_rtr_pref is enabled.
901 -1 if accept_ra_rtr_pref is disabled.
903 accept_ra_rtr_pref - BOOLEAN
904 Accept Router Preference in RA.
906 Functional default: enabled if accept_ra is enabled.
907 disabled if accept_ra is disabled.
909 accept_redirects - BOOLEAN
912 Functional default: enabled if local forwarding is disabled.
913 disabled if local forwarding is enabled.
915 accept_source_route - INTEGER
916 Accept source routing (routing extension header).
918 >= 0: Accept only routing header type 2.
919 < 0: Do not accept routing header.
924 Autoconfigure addresses using Prefix Information in Router
927 Functional default: enabled if accept_ra_pinfo is enabled.
928 disabled if accept_ra_pinfo is disabled.
930 dad_transmits - INTEGER
931 The amount of Duplicate Address Detection probes to send.
935 Configure interface-specific Host/Router behaviour.
937 Note: It is recommended to have the same setting on all
938 interfaces; mixed router/host scenarios are rather uncommon.
942 By default, Host behaviour is assumed. This means:
944 1. IsRouter flag is not set in Neighbour Advertisements.
945 2. Router Solicitations are being sent when necessary.
946 3. If accept_ra is TRUE (default), accept Router
947 Advertisements (and do autoconfiguration).
948 4. If accept_redirects is TRUE (default), accept Redirects.
952 If local forwarding is enabled, Router behaviour is assumed.
953 This means exactly the reverse from the above:
955 1. IsRouter flag is set in Neighbour Advertisements.
956 2. Router Solicitations are not sent.
957 3. Router Advertisements are ignored.
958 4. Redirects are ignored.
960 Default: FALSE if global forwarding is disabled (default),
964 Default Hop Limit to set.
968 Default Maximum Transfer Unit
969 Default: 1280 (IPv6 required minimum)
971 router_probe_interval - INTEGER
972 Minimum interval (in seconds) between Router Probing described
977 router_solicitation_delay - INTEGER
978 Number of seconds to wait after interface is brought up
979 before sending Router Solicitations.
982 router_solicitation_interval - INTEGER
983 Number of seconds to wait between Router Solicitations.
986 router_solicitations - INTEGER
987 Number of Router Solicitations to send until assuming no
991 use_tempaddr - INTEGER
992 Preference for Privacy Extensions (RFC3041).
993 <= 0 : disable Privacy Extensions
994 == 1 : enable Privacy Extensions, but prefer public
995 addresses over temporary addresses.
996 > 1 : enable Privacy Extensions and prefer temporary
997 addresses over public addresses.
998 Default: 0 (for most devices)
999 -1 (for point-to-point devices and loopback devices)
1001 temp_valid_lft - INTEGER
1002 valid lifetime (in seconds) for temporary addresses.
1003 Default: 604800 (7 days)
1005 temp_prefered_lft - INTEGER
1006 Preferred lifetime (in seconds) for temporary addresses.
1007 Default: 86400 (1 day)
1009 max_desync_factor - INTEGER
1010 Maximum value for DESYNC_FACTOR, which is a random value
1011 that ensures that clients don't synchronize with each
1012 other and generate new addresses at exactly the same time.
1013 value is in seconds.
1016 regen_max_retry - INTEGER
1017 Number of attempts before give up attempting to generate
1018 valid temporary addresses.
1021 max_addresses - INTEGER
1022 Number of maximum addresses per interface. 0 disables limitation.
1023 It is recommended not set too large value (or 0) because it would
1024 be too easy way to crash kernel to allow to create too much of
1025 autoconfigured addresses.
1028 disable_ipv6 - BOOLEAN
1029 Disable IPv6 operation.
1030 Default: FALSE (enable IPv6 operation)
1034 Limit the maximal rates for sending ICMPv6 packets.
1035 0 to disable any limiting,
1036 otherwise the minimal space between responses in milliseconds.
1041 Pekka Savola <pekkas@netcore.fi>
1042 YOSHIFUJI Hideaki / USAGI Project <yoshfuji@linux-ipv6.org>
1045 /proc/sys/net/bridge/* Variables:
1047 bridge-nf-call-arptables - BOOLEAN
1048 1 : pass bridged ARP traffic to arptables' FORWARD chain.
1052 bridge-nf-call-iptables - BOOLEAN
1053 1 : pass bridged IPv4 traffic to iptables' chains.
1057 bridge-nf-call-ip6tables - BOOLEAN
1058 1 : pass bridged IPv6 traffic to ip6tables' chains.
1062 bridge-nf-filter-vlan-tagged - BOOLEAN
1063 1 : pass bridged vlan-tagged ARP/IP/IPv6 traffic to {arp,ip,ip6}tables.
1067 bridge-nf-filter-pppoe-tagged - BOOLEAN
1068 1 : pass bridged pppoe-tagged IP/IPv6 traffic to {ip,ip6}tables.
1076 discovery_slots FIXME
1077 discovery_timeout FIXME
1078 fast_poll_increase FIXME
1079 ip6_queue_maxlen FIXME
1080 lap_keepalive_time FIXME
1083 max_dgram_qlen FIXME
1084 max_noreply_time FIXME
1085 max_tx_data_size FIXME
1087 min_tx_turn_time FIXME
1090 no_cong_thresh FIXME
1092 warn_noreply_time FIXME