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 - BOOLEAN
19 Disable Path MTU Discovery.
23 default 552 - minimum discovered Path MTU
25 route/max_size - INTEGER
26 Maximum number of routes allowed in the kernel. Increase
27 this when using large numbers of interfaces and/or routes.
29 neigh/default/gc_thresh3 - INTEGER
30 Maximum number of neighbor entries allowed. Increase this
31 when using large numbers of interfaces and when communicating
32 with large numbers of directly-connected peers.
35 Time, in seconds, that cached PMTU information is kept.
38 The advertised MSS depends on the first hop route MTU, but will
39 never be lower than this setting.
41 rt_cache_rebuild_count - INTEGER
42 The per net-namespace route cache emergency rebuild threshold.
43 Any net-namespace having its route cache rebuilt due to
44 a hash bucket chain being too long more than this many times
45 will have its route caching disabled
49 ipfrag_high_thresh - INTEGER
50 Maximum memory used to reassemble IP fragments. When
51 ipfrag_high_thresh bytes of memory is allocated for this purpose,
52 the fragment handler will toss packets until ipfrag_low_thresh
55 ipfrag_low_thresh - INTEGER
56 See ipfrag_high_thresh
59 Time in seconds to keep an IP fragment in memory.
61 ipfrag_secret_interval - INTEGER
62 Regeneration interval (in seconds) of the hash secret (or lifetime
63 for the hash secret) for IP fragments.
66 ipfrag_max_dist - INTEGER
67 ipfrag_max_dist is a non-negative integer value which defines the
68 maximum "disorder" which is allowed among fragments which share a
69 common IP source address. Note that reordering of packets is
70 not unusual, but if a large number of fragments arrive from a source
71 IP address while a particular fragment queue remains incomplete, it
72 probably indicates that one or more fragments belonging to that queue
73 have been lost. When ipfrag_max_dist is positive, an additional check
74 is done on fragments before they are added to a reassembly queue - if
75 ipfrag_max_dist (or more) fragments have arrived from a particular IP
76 address between additions to any IP fragment queue using that source
77 address, it's presumed that one or more fragments in the queue are
78 lost. The existing fragment queue will be dropped, and a new one
79 started. An ipfrag_max_dist value of zero disables this check.
81 Using a very small value, e.g. 1 or 2, for ipfrag_max_dist can
82 result in unnecessarily dropping fragment queues when normal
83 reordering of packets occurs, which could lead to poor application
84 performance. Using a very large value, e.g. 50000, increases the
85 likelihood of incorrectly reassembling IP fragments that originate
86 from different IP datagrams, which could result in data corruption.
91 inet_peer_threshold - INTEGER
92 The approximate size of the storage. Starting from this threshold
93 entries will be thrown aggressively. This threshold also determines
94 entries' time-to-live and time intervals between garbage collection
95 passes. More entries, less time-to-live, less GC interval.
97 inet_peer_minttl - INTEGER
98 Minimum time-to-live of entries. Should be enough to cover fragment
99 time-to-live on the reassembling side. This minimum time-to-live is
100 guaranteed if the pool size is less than inet_peer_threshold.
103 inet_peer_maxttl - INTEGER
104 Maximum time-to-live of entries. Unused entries will expire after
105 this period of time if there is no memory pressure on the pool (i.e.
106 when the number of entries in the pool is very small).
112 Limit of socket listen() backlog, known in userspace as SOMAXCONN.
113 Defaults to 128. See also tcp_max_syn_backlog for additional tuning
117 Controls Appropriate Byte Count (ABC) defined in RFC3465.
118 ABC is a way of increasing congestion window (cwnd) more slowly
119 in response to partial acknowledgments.
121 0 increase cwnd once per acknowledgment (no ABC)
122 1 increase cwnd once per acknowledgment of full sized segment
123 2 allow increase cwnd by two if acknowledgment is
124 of two segments to compensate for delayed acknowledgments.
127 tcp_abort_on_overflow - BOOLEAN
128 If listening service is too slow to accept new connections,
129 reset them. Default state is FALSE. It means that if overflow
130 occurred due to a burst, connection will recover. Enable this
131 option _only_ if you are really sure that listening daemon
132 cannot be tuned to accept connections faster. Enabling this
133 option can harm clients of your server.
135 tcp_adv_win_scale - INTEGER
136 Count buffering overhead as bytes/2^tcp_adv_win_scale
137 (if tcp_adv_win_scale > 0) or bytes-bytes/2^(-tcp_adv_win_scale),
139 Possible values are [-31, 31], inclusive.
142 tcp_allowed_congestion_control - STRING
143 Show/set the congestion control choices available to non-privileged
144 processes. The list is a subset of those listed in
145 tcp_available_congestion_control.
146 Default is "reno" and the default setting (tcp_congestion_control).
148 tcp_app_win - INTEGER
149 Reserve max(window/2^tcp_app_win, mss) of window for application
150 buffer. Value 0 is special, it means that nothing is reserved.
153 tcp_available_congestion_control - STRING
154 Shows the available congestion control choices that are registered.
155 More congestion control algorithms may be available as modules,
158 tcp_base_mss - INTEGER
159 The initial value of search_low to be used by the packetization layer
160 Path MTU discovery (MTU probing). If MTU probing is enabled,
161 this is the initial MSS used by the connection.
163 tcp_min_snd_mss - INTEGER
164 TCP SYN and SYNACK messages usually advertise an ADVMSS option,
165 as described in RFC 1122 and RFC 6691.
166 If this ADVMSS option is smaller than tcp_min_snd_mss,
167 it is silently capped to tcp_min_snd_mss.
169 Default : 48 (at least 8 bytes of payload per segment)
171 tcp_congestion_control - STRING
172 Set the congestion control algorithm to be used for new
173 connections. The algorithm "reno" is always available, but
174 additional choices may be available based on kernel configuration.
175 Default is set as part of kernel configuration.
177 tcp_cookie_size - INTEGER
178 Default size of TCP Cookie Transactions (TCPCT) option, that may be
179 overridden on a per socket basis by the TCPCT socket option.
180 Values greater than the maximum (16) are interpreted as the maximum.
181 Values greater than zero and less than the minimum (8) are interpreted
182 as the minimum. Odd values are interpreted as the next even value.
186 Allows TCP to send "duplicate" SACKs.
189 Enable Explicit Congestion Notification (ECN) in TCP. ECN is only
190 used when both ends of the TCP flow support it. It is useful to
191 avoid losses due to congestion (when the bottleneck router supports
196 2 Only server-side ECN enabled. If the other end does
197 not support ECN, behavior is like with ECN disabled.
201 Enable FACK congestion avoidance and fast retransmission.
202 The value is not used, if tcp_sack is not enabled.
204 tcp_fin_timeout - INTEGER
205 Time to hold socket in state FIN-WAIT-2, if it was closed
206 by our side. Peer can be broken and never close its side,
207 or even died unexpectedly. Default value is 60sec.
208 Usual value used in 2.2 was 180 seconds, you may restore
209 it, but remember that if your machine is even underloaded WEB server,
210 you risk to overflow memory with kilotons of dead sockets,
211 FIN-WAIT-2 sockets are less dangerous than FIN-WAIT-1,
212 because they eat maximum 1.5K of memory, but they tend
213 to live longer. Cf. tcp_max_orphans.
216 Enables Forward RTO-Recovery (F-RTO) defined in RFC4138.
217 F-RTO is an enhanced recovery algorithm for TCP retransmission
218 timeouts. It is particularly beneficial in wireless environments
219 where packet loss is typically due to random radio interference
220 rather than intermediate router congestion. F-RTO is sender-side
221 only modification. Therefore it does not require any support from
224 If set to 1, basic version is enabled. 2 enables SACK enhanced
225 F-RTO if flow uses SACK. The basic version can be used also when
226 SACK is in use though scenario(s) with it exists where F-RTO
227 interacts badly with the packet counting of the SACK enabled TCP
230 tcp_frto_response - INTEGER
231 When F-RTO has detected that a TCP retransmission timeout was
232 spurious (i.e, the timeout would have been avoided had TCP set a
233 longer retransmission timeout), TCP has several options what to do
234 next. Possible values are:
235 0 Rate halving based; a smooth and conservative response,
236 results in halved cwnd and ssthresh after one RTT
237 1 Very conservative response; not recommended because even
238 though being valid, it interacts poorly with the rest of
239 Linux TCP, halves cwnd and ssthresh immediately
240 2 Aggressive response; undoes congestion control measures
241 that are now known to be unnecessary (ignoring the
242 possibility of a lost retransmission that would require
243 TCP to be more cautious), cwnd and ssthresh are restored
244 to the values prior timeout
245 Default: 0 (rate halving based)
247 tcp_keepalive_time - INTEGER
248 How often TCP sends out keepalive messages when keepalive is enabled.
251 tcp_keepalive_probes - INTEGER
252 How many keepalive probes TCP sends out, until it decides that the
253 connection is broken. Default value: 9.
255 tcp_keepalive_intvl - INTEGER
256 How frequently the probes are send out. Multiplied by
257 tcp_keepalive_probes it is time to kill not responding connection,
258 after probes started. Default value: 75sec i.e. connection
259 will be aborted after ~11 minutes of retries.
261 tcp_low_latency - BOOLEAN
262 If set, the TCP stack makes decisions that prefer lower
263 latency as opposed to higher throughput. By default, this
264 option is not set meaning that higher throughput is preferred.
265 An example of an application where this default should be
266 changed would be a Beowulf compute cluster.
269 tcp_max_orphans - INTEGER
270 Maximal number of TCP sockets not attached to any user file handle,
271 held by system. If this number is exceeded orphaned connections are
272 reset immediately and warning is printed. This limit exists
273 only to prevent simple DoS attacks, you _must_ not rely on this
274 or lower the limit artificially, but rather increase it
275 (probably, after increasing installed memory),
276 if network conditions require more than default value,
277 and tune network services to linger and kill such states
278 more aggressively. Let me to remind again: each orphan eats
279 up to ~64K of unswappable memory.
281 tcp_max_ssthresh - INTEGER
282 Limited Slow-Start for TCP with large congestion windows (cwnd) defined in
283 RFC3742. Limited slow-start is a mechanism to limit growth of the cwnd
284 on the region where cwnd is larger than tcp_max_ssthresh. TCP increases cwnd
285 by at most tcp_max_ssthresh segments, and by at least tcp_max_ssthresh/2
286 segments per RTT when the cwnd is above tcp_max_ssthresh.
287 If TCP connection increased cwnd to thousands (or tens of thousands) segments,
288 and thousands of packets were being dropped during slow-start, you can set
289 tcp_max_ssthresh to improve performance for new TCP connection.
292 tcp_max_syn_backlog - INTEGER
293 Maximal number of remembered connection requests, which have not
294 received an acknowledgment from connecting client.
295 The minimal value is 128 for low memory machines, and it will
296 increase in proportion to the memory of machine.
297 If server suffers from overload, try increasing this number.
299 tcp_max_tw_buckets - INTEGER
300 Maximal number of timewait sockets held by system simultaneously.
301 If this number is exceeded time-wait socket is immediately destroyed
302 and warning is printed. This limit exists only to prevent
303 simple DoS attacks, you _must_ not lower the limit artificially,
304 but rather increase it (probably, after increasing installed memory),
305 if network conditions require more than default value.
307 tcp_mem - vector of 3 INTEGERs: min, pressure, max
308 min: below this number of pages TCP is not bothered about its
311 pressure: when amount of memory allocated by TCP exceeds this number
312 of pages, TCP moderates its memory consumption and enters memory
313 pressure mode, which is exited when memory consumption falls
316 max: number of pages allowed for queueing by all TCP sockets.
318 Defaults are calculated at boot time from amount of available
321 tcp_moderate_rcvbuf - BOOLEAN
322 If set, TCP performs receive buffer auto-tuning, attempting to
323 automatically size the buffer (no greater than tcp_rmem[2]) to
324 match the size required by the path for full throughput. Enabled by
327 tcp_mtu_probing - INTEGER
328 Controls TCP Packetization-Layer Path MTU Discovery. Takes three
331 1 - Disabled by default, enabled when an ICMP black hole detected
332 2 - Always enabled, use initial MSS of tcp_base_mss.
334 tcp_no_metrics_save - BOOLEAN
335 By default, TCP saves various connection metrics in the route cache
336 when the connection closes, so that connections established in the
337 near future can use these to set initial conditions. Usually, this
338 increases overall performance, but may sometimes cause performance
339 degradation. If set, TCP will not cache metrics on closing
342 tcp_orphan_retries - INTEGER
343 This value influences the timeout of a locally closed TCP connection,
344 when RTO retransmissions remain unacknowledged.
345 See tcp_retries2 for more details.
347 The default value is 8.
348 If your machine is a loaded WEB server,
349 you should think about lowering this value, such sockets
350 may consume significant resources. Cf. tcp_max_orphans.
352 tcp_reordering - INTEGER
353 Maximal reordering of packets in a TCP stream.
356 tcp_retrans_collapse - BOOLEAN
357 Bug-to-bug compatibility with some broken printers.
358 On retransmit try to send bigger packets to work around bugs in
361 tcp_retries1 - INTEGER
362 This value influences the time, after which TCP decides, that
363 something is wrong due to unacknowledged RTO retransmissions,
364 and reports this suspicion to the network layer.
365 See tcp_retries2 for more details.
367 RFC 1122 recommends at least 3 retransmissions, which is the
370 tcp_retries2 - INTEGER
371 This value influences the timeout of an alive TCP connection,
372 when RTO retransmissions remain unacknowledged.
373 Given a value of N, a hypothetical TCP connection following
374 exponential backoff with an initial RTO of TCP_RTO_MIN would
375 retransmit N times before killing the connection at the (N+1)th RTO.
377 The default value of 15 yields a hypothetical timeout of 924.6
378 seconds and is a lower bound for the effective timeout.
379 TCP will effectively time out at the first RTO which exceeds the
380 hypothetical timeout.
382 RFC 1122 recommends at least 100 seconds for the timeout,
383 which corresponds to a value of at least 8.
385 tcp_rfc1337 - BOOLEAN
386 If set, the TCP stack behaves conforming to RFC1337. If unset,
387 we are not conforming to RFC, but prevent TCP TIME_WAIT
391 tcp_rmem - vector of 3 INTEGERs: min, default, max
392 min: Minimal size of receive buffer used by TCP sockets.
393 It is guaranteed to each TCP socket, even under moderate memory
397 default: initial size of receive buffer used by TCP sockets.
398 This value overrides net.core.rmem_default used by other protocols.
399 Default: 87380 bytes. This value results in window of 65535 with
400 default setting of tcp_adv_win_scale and tcp_app_win:0 and a bit
401 less for default tcp_app_win. See below about these variables.
403 max: maximal size of receive buffer allowed for automatically
404 selected receiver buffers for TCP socket. This value does not override
405 net.core.rmem_max. Calling setsockopt() with SO_RCVBUF disables
406 automatic tuning of that socket's receive buffer size, in which
407 case this value is ignored.
408 Default: between 87380B and 6MB, depending on RAM size.
411 Enable select acknowledgments (SACKS).
413 tcp_slow_start_after_idle - BOOLEAN
414 If set, provide RFC2861 behavior and time out the congestion
415 window after an idle period. An idle period is defined at
416 the current RTO. If unset, the congestion window will not
417 be timed out after an idle period.
421 Use the Host requirements interpretation of the TCP urgent pointer field.
422 Most hosts use the older BSD interpretation, so if you turn this on
423 Linux might not communicate correctly with them.
426 tcp_synack_retries - INTEGER
427 Number of times SYNACKs for a passive TCP connection attempt will
428 be retransmitted. Should not be higher than 255. Default value
429 is 5, which corresponds to ~180seconds.
431 tcp_syncookies - BOOLEAN
432 Only valid when the kernel was compiled with CONFIG_SYNCOOKIES
433 Send out syncookies when the syn backlog queue of a socket
434 overflows. This is to prevent against the common 'SYN flood attack'
437 Note, that syncookies is fallback facility.
438 It MUST NOT be used to help highly loaded servers to stand
439 against legal connection rate. If you see SYN flood warnings
440 in your logs, but investigation shows that they occur
441 because of overload with legal connections, you should tune
442 another parameters until this warning disappear.
443 See: tcp_max_syn_backlog, tcp_synack_retries, tcp_abort_on_overflow.
445 syncookies seriously violate TCP protocol, do not allow
446 to use TCP extensions, can result in serious degradation
447 of some services (f.e. SMTP relaying), visible not by you,
448 but your clients and relays, contacting you. While you see
449 SYN flood warnings in logs not being really flooded, your server
450 is seriously misconfigured.
452 tcp_syn_retries - INTEGER
453 Number of times initial SYNs for an active TCP connection attempt
454 will be retransmitted. Should not be higher than 255. Default value
455 is 5, which corresponds to ~180seconds.
457 tcp_timestamps - BOOLEAN
458 Enable timestamps as defined in RFC1323.
460 tcp_tso_win_divisor - INTEGER
461 This allows control over what percentage of the congestion window
462 can be consumed by a single TSO frame.
463 The setting of this parameter is a choice between burstiness and
464 building larger TSO frames.
467 tcp_tw_recycle - BOOLEAN
468 Enable fast recycling TIME-WAIT sockets. Default value is 0.
469 It should not be changed without advice/request of technical
472 tcp_tw_reuse - BOOLEAN
473 Allow to reuse TIME-WAIT sockets for new connections when it is
474 safe from protocol viewpoint. Default value is 0.
475 It should not be changed without advice/request of technical
478 tcp_window_scaling - BOOLEAN
479 Enable window scaling as defined in RFC1323.
481 tcp_wmem - vector of 3 INTEGERs: min, default, max
482 min: Amount of memory reserved for send buffers for TCP sockets.
483 Each TCP socket has rights to use it due to fact of its birth.
486 default: initial size of send buffer used by TCP sockets. This
487 value overrides net.core.wmem_default used by other protocols.
488 It is usually lower than net.core.wmem_default.
491 max: Maximal amount of memory allowed for automatically tuned
492 send buffers for TCP sockets. This value does not override
493 net.core.wmem_max. Calling setsockopt() with SO_SNDBUF disables
494 automatic tuning of that socket's send buffer size, in which case
495 this value is ignored.
496 Default: between 64K and 4MB, depending on RAM size.
498 tcp_workaround_signed_windows - BOOLEAN
499 If set, assume no receipt of a window scaling option means the
500 remote TCP is broken and treats the window as a signed quantity.
501 If unset, assume the remote TCP is not broken even if we do
502 not receive a window scaling option from them.
505 tcp_dma_copybreak - INTEGER
506 Lower limit, in bytes, of the size of socket reads that will be
507 offloaded to a DMA copy engine, if one is present in the system
508 and CONFIG_NET_DMA is enabled.
511 tcp_thin_linear_timeouts - BOOLEAN
512 Enable dynamic triggering of linear timeouts for thin streams.
513 If set, a check is performed upon retransmission by timeout to
514 determine if the stream is thin (less than 4 packets in flight).
515 As long as the stream is found to be thin, up to 6 linear
516 timeouts may be performed before exponential backoff mode is
517 initiated. This improves retransmission latency for
518 non-aggressive thin streams, often found to be time-dependent.
519 For more information on thin streams, see
520 Documentation/networking/tcp-thin.txt
523 tcp_thin_dupack - BOOLEAN
524 Enable dynamic triggering of retransmissions after one dupACK
525 for thin streams. If set, a check is performed upon reception
526 of a dupACK to determine if the stream is thin (less than 4
527 packets in flight). As long as the stream is found to be thin,
528 data is retransmitted on the first received dupACK. This
529 improves retransmission latency for non-aggressive thin
530 streams, often found to be time-dependent.
531 For more information on thin streams, see
532 Documentation/networking/tcp-thin.txt
535 tcp_challenge_ack_limit - INTEGER
536 Limits number of Challenge ACK sent per second, as recommended
537 in RFC 5961 (Improving TCP's Robustness to Blind In-Window Attacks)
542 udp_mem - vector of 3 INTEGERs: min, pressure, max
543 Number of pages allowed for queueing by all UDP sockets.
545 min: Below this number of pages UDP is not bothered about its
546 memory appetite. When amount of memory allocated by UDP exceeds
547 this number, UDP starts to moderate memory usage.
549 pressure: This value was introduced to follow format of tcp_mem.
551 max: Number of pages allowed for queueing by all UDP sockets.
553 Default is calculated at boot time from amount of available memory.
555 udp_rmem_min - INTEGER
556 Minimal size of receive buffer used by UDP sockets in moderation.
557 Each UDP socket is able to use the size for receiving data, even if
558 total pages of UDP sockets exceed udp_mem pressure. The unit is byte.
561 udp_wmem_min - INTEGER
562 Minimal size of send buffer used by UDP sockets in moderation.
563 Each UDP socket is able to use the size for sending data, even if
564 total pages of UDP sockets exceed udp_mem pressure. The unit is byte.
569 cipso_cache_enable - BOOLEAN
570 If set, enable additions to and lookups from the CIPSO label mapping
571 cache. If unset, additions are ignored and lookups always result in a
572 miss. However, regardless of the setting the cache is still
573 invalidated when required when means you can safely toggle this on and
574 off and the cache will always be "safe".
577 cipso_cache_bucket_size - INTEGER
578 The CIPSO label cache consists of a fixed size hash table with each
579 hash bucket containing a number of cache entries. This variable limits
580 the number of entries in each hash bucket; the larger the value the
581 more CIPSO label mappings that can be cached. When the number of
582 entries in a given hash bucket reaches this limit adding new entries
583 causes the oldest entry in the bucket to be removed to make room.
586 cipso_rbm_optfmt - BOOLEAN
587 Enable the "Optimized Tag 1 Format" as defined in section 3.4.2.6 of
588 the CIPSO draft specification (see Documentation/netlabel for details).
589 This means that when set the CIPSO tag will be padded with empty
590 categories in order to make the packet data 32-bit aligned.
593 cipso_rbm_structvalid - BOOLEAN
594 If set, do a very strict check of the CIPSO option when
595 ip_options_compile() is called. If unset, relax the checks done during
596 ip_options_compile(). Either way is "safe" as errors are caught else
597 where in the CIPSO processing code but setting this to 0 (False) should
598 result in less work (i.e. it should be faster) but could cause problems
599 with other implementations that require strict checking.
604 ip_local_port_range - 2 INTEGERS
605 Defines the local port range that is used by TCP and UDP to
606 choose the local port. The first number is the first, the
607 second the last local port number. Default value depends on
608 amount of memory available on the system:
610 < 128Mb 1024-4999 or even less.
611 This number defines number of active connections, which this
612 system can issue simultaneously to systems not supporting
613 TCP extensions (timestamps). With tcp_tw_recycle enabled
614 (i.e. by default) range 1024-4999 is enough to issue up to
615 2000 connections per second to systems supporting timestamps.
617 ip_local_reserved_ports - list of comma separated ranges
618 Specify the ports which are reserved for known third-party
619 applications. These ports will not be used by automatic port
620 assignments (e.g. when calling connect() or bind() with port
621 number 0). Explicit port allocation behavior is unchanged.
623 The format used for both input and output is a comma separated
624 list of ranges (e.g. "1,2-4,10-10" for ports 1, 2, 3, 4 and
625 10). Writing to the file will clear all previously reserved
626 ports and update the current list with the one given in the
629 Note that ip_local_port_range and ip_local_reserved_ports
630 settings are independent and both are considered by the kernel
631 when determining which ports are available for automatic port
634 You can reserve ports which are not in the current
635 ip_local_port_range, e.g.:
637 $ cat /proc/sys/net/ipv4/ip_local_port_range
639 $ cat /proc/sys/net/ipv4/ip_local_reserved_ports
642 although this is redundant. However such a setting is useful
643 if later the port range is changed to a value that will
644 include the reserved ports.
648 ip_nonlocal_bind - BOOLEAN
649 If set, allows processes to bind() to non-local IP addresses,
650 which can be quite useful - but may break some applications.
654 If set non-zero, enables support for dynamic addresses.
655 If set to a non-zero value larger than 1, a kernel log
656 message will be printed when dynamic address rewriting
660 icmp_echo_ignore_all - BOOLEAN
661 If set non-zero, then the kernel will ignore all ICMP ECHO
665 icmp_echo_ignore_broadcasts - BOOLEAN
666 If set non-zero, then the kernel will ignore all ICMP ECHO and
667 TIMESTAMP requests sent to it via broadcast/multicast.
670 icmp_ratelimit - INTEGER
671 Limit the maximal rates for sending ICMP packets whose type matches
672 icmp_ratemask (see below) to specific targets.
673 0 to disable any limiting,
674 otherwise the minimal space between responses in milliseconds.
677 icmp_ratemask - INTEGER
678 Mask made of ICMP types for which rates are being limited.
679 Significant bits: IHGFEDCBA9876543210
680 Default mask: 0000001100000011000 (6168)
682 Bit definitions (see include/linux/icmp.h):
684 3 Destination Unreachable *
689 C Parameter Problem *
694 H Address Mask Request
697 * These are rate limited by default (see default mask above)
699 icmp_ignore_bogus_error_responses - BOOLEAN
700 Some routers violate RFC1122 by sending bogus responses to broadcast
701 frames. Such violations are normally logged via a kernel warning.
702 If this is set to TRUE, the kernel will not give such warnings, which
703 will avoid log file clutter.
706 icmp_errors_use_inbound_ifaddr - BOOLEAN
708 If zero, icmp error messages are sent with the primary address of
709 the exiting interface.
711 If non-zero, the message will be sent with the primary address of
712 the interface that received the packet that caused the icmp error.
713 This is the behaviour network many administrators will expect from
714 a router. And it can make debugging complicated network layouts
717 Note that if no primary address exists for the interface selected,
718 then the primary address of the first non-loopback interface that
719 has one will be used regardless of this setting.
723 igmp_max_memberships - INTEGER
724 Change the maximum number of multicast groups we can subscribe to.
727 Theoretical maximum value is bounded by having to send a membership
728 report in a single datagram (i.e. the report can't span multiple
729 datagrams, or risk confusing the switch and leaving groups you don't
732 The number of supported groups 'M' is bounded by the number of group
733 report entries you can fit into a single datagram of 65535 bytes.
735 M = 65536-sizeof (ip header)/(sizeof(Group record))
737 Group records are variable length, with a minimum of 12 bytes.
738 So net.ipv4.igmp_max_memberships should not be set higher than:
740 (65536-24) / 12 = 5459
742 The value 5459 assumes no IP header options, so in practice
743 this number may be lower.
745 conf/interface/* changes special settings per interface (where
746 "interface" is the name of your network interface)
748 conf/all/* is special, changes the settings for all interfaces
750 log_martians - BOOLEAN
751 Log packets with impossible addresses to kernel log.
752 log_martians for the interface will be enabled if at least one of
753 conf/{all,interface}/log_martians is set to TRUE,
754 it will be disabled otherwise
756 accept_redirects - BOOLEAN
757 Accept ICMP redirect messages.
758 accept_redirects for the interface will be enabled if:
759 - both conf/{all,interface}/accept_redirects are TRUE in the case
760 forwarding for the interface is enabled
762 - at least one of conf/{all,interface}/accept_redirects is TRUE in the
763 case forwarding for the interface is disabled
764 accept_redirects for the interface will be disabled otherwise
769 Enable IP forwarding on this interface.
771 mc_forwarding - BOOLEAN
772 Do multicast routing. The kernel needs to be compiled with CONFIG_MROUTE
773 and a multicast routing daemon is required.
774 conf/all/mc_forwarding must also be set to TRUE to enable multicast
775 routing for the interface
778 Integer value used to differentiate the devices by the medium they
779 are attached to. Two devices can have different id values when
780 the broadcast packets are received only on one of them.
781 The default value 0 means that the device is the only interface
782 to its medium, value of -1 means that medium is not known.
784 Currently, it is used to change the proxy_arp behavior:
785 the proxy_arp feature is enabled for packets forwarded between
786 two devices attached to different media.
790 proxy_arp for the interface will be enabled if at least one of
791 conf/{all,interface}/proxy_arp is set to TRUE,
792 it will be disabled otherwise
794 proxy_arp_pvlan - BOOLEAN
795 Private VLAN proxy arp.
796 Basically allow proxy arp replies back to the same interface
797 (from which the ARP request/solicitation was received).
799 This is done to support (ethernet) switch features, like RFC
800 3069, where the individual ports are NOT allowed to
801 communicate with each other, but they are allowed to talk to
802 the upstream router. As described in RFC 3069, it is possible
803 to allow these hosts to communicate through the upstream
804 router by proxy_arp'ing. Don't need to be used together with
807 This technology is known by different names:
808 In RFC 3069 it is called VLAN Aggregation.
809 Cisco and Allied Telesyn call it Private VLAN.
810 Hewlett-Packard call it Source-Port filtering or port-isolation.
811 Ericsson call it MAC-Forced Forwarding (RFC Draft).
813 shared_media - BOOLEAN
814 Send(router) or accept(host) RFC1620 shared media redirects.
815 Overrides ip_secure_redirects.
816 shared_media for the interface will be enabled if at least one of
817 conf/{all,interface}/shared_media is set to TRUE,
818 it will be disabled otherwise
821 secure_redirects - BOOLEAN
822 Accept ICMP redirect messages only for gateways,
823 listed in default gateway list.
824 secure_redirects for the interface will be enabled if at least one of
825 conf/{all,interface}/secure_redirects is set to TRUE,
826 it will be disabled otherwise
829 send_redirects - BOOLEAN
830 Send redirects, if router.
831 send_redirects for the interface will be enabled if at least one of
832 conf/{all,interface}/send_redirects is set to TRUE,
833 it will be disabled otherwise
836 bootp_relay - BOOLEAN
837 Accept packets with source address 0.b.c.d destined
838 not to this host as local ones. It is supposed, that
839 BOOTP relay daemon will catch and forward such packets.
840 conf/all/bootp_relay must also be set to TRUE to enable BOOTP relay
845 accept_source_route - BOOLEAN
846 Accept packets with SRR option.
847 conf/all/accept_source_route must also be set to TRUE to accept packets
848 with SRR option on the interface
849 default TRUE (router)
852 accept_local - BOOLEAN
853 Accept packets with local source addresses. In combination with
854 suitable routing, this can be used to direct packets between two
855 local interfaces over the wire and have them accepted properly.
859 0 - No source validation.
860 1 - Strict mode as defined in RFC3704 Strict Reverse Path
861 Each incoming packet is tested against the FIB and if the interface
862 is not the best reverse path the packet check will fail.
863 By default failed packets are discarded.
864 2 - Loose mode as defined in RFC3704 Loose Reverse Path
865 Each incoming packet's source address is also tested against the FIB
866 and if the source address is not reachable via any interface
867 the packet check will fail.
869 Current recommended practice in RFC3704 is to enable strict mode
870 to prevent IP spoofing from DDos attacks. If using asymmetric routing
871 or other complicated routing, then loose mode is recommended.
873 The max value from conf/{all,interface}/rp_filter is used
874 when doing source validation on the {interface}.
876 Default value is 0. Note that some distributions enable it
880 1 - Allows you to have multiple network interfaces on the same
881 subnet, and have the ARPs for each interface be answered
882 based on whether or not the kernel would route a packet from
883 the ARP'd IP out that interface (therefore you must use source
884 based routing for this to work). In other words it allows control
885 of which cards (usually 1) will respond to an arp request.
887 0 - (default) The kernel can respond to arp requests with addresses
888 from other interfaces. This may seem wrong but it usually makes
889 sense, because it increases the chance of successful communication.
890 IP addresses are owned by the complete host on Linux, not by
891 particular interfaces. Only for more complex setups like load-
892 balancing, does this behaviour cause problems.
894 arp_filter for the interface will be enabled if at least one of
895 conf/{all,interface}/arp_filter is set to TRUE,
896 it will be disabled otherwise
898 arp_announce - INTEGER
899 Define different restriction levels for announcing the local
900 source IP address from IP packets in ARP requests sent on
902 0 - (default) Use any local address, configured on any interface
903 1 - Try to avoid local addresses that are not in the target's
904 subnet for this interface. This mode is useful when target
905 hosts reachable via this interface require the source IP
906 address in ARP requests to be part of their logical network
907 configured on the receiving interface. When we generate the
908 request we will check all our subnets that include the
909 target IP and will preserve the source address if it is from
910 such subnet. If there is no such subnet we select source
911 address according to the rules for level 2.
912 2 - Always use the best local address for this target.
913 In this mode we ignore the source address in the IP packet
914 and try to select local address that we prefer for talks with
915 the target host. Such local address is selected by looking
916 for primary IP addresses on all our subnets on the outgoing
917 interface that include the target IP address. If no suitable
918 local address is found we select the first local address
919 we have on the outgoing interface or on all other interfaces,
920 with the hope we will receive reply for our request and
921 even sometimes no matter the source IP address we announce.
923 The max value from conf/{all,interface}/arp_announce is used.
925 Increasing the restriction level gives more chance for
926 receiving answer from the resolved target while decreasing
927 the level announces more valid sender's information.
930 Define different modes for sending replies in response to
931 received ARP requests that resolve local target IP addresses:
932 0 - (default): reply for any local target IP address, configured
934 1 - reply only if the target IP address is local address
935 configured on the incoming interface
936 2 - reply only if the target IP address is local address
937 configured on the incoming interface and both with the
938 sender's IP address are part from same subnet on this interface
939 3 - do not reply for local addresses configured with scope host,
940 only resolutions for global and link addresses are replied
942 8 - do not reply for all local addresses
944 The max value from conf/{all,interface}/arp_ignore is used
945 when ARP request is received on the {interface}
948 Define mode for notification of address and device changes.
949 0 - (default): do nothing
950 1 - Generate gratuitous arp requests when device is brought up
951 or hardware address changes.
954 Define behavior for gratuitous ARP frames who's IP is not
955 already present in the ARP table:
956 0 - don't create new entries in the ARP table
957 1 - create new entries in the ARP table
959 Both replies and requests type gratuitous arp will trigger the
960 ARP table to be updated, if this setting is on.
962 If the ARP table already contains the IP address of the
963 gratuitous arp frame, the arp table will be updated regardless
964 if this setting is on or off.
967 app_solicit - INTEGER
968 The maximum number of probes to send to the user space ARP daemon
969 via netlink before dropping back to multicast probes (see
970 mcast_solicit). Defaults to 0.
972 disable_policy - BOOLEAN
973 Disable IPSEC policy (SPD) for this interface
975 disable_xfrm - BOOLEAN
976 Disable IPSEC encryption on this interface, whatever the policy
981 Allows you to write a number, which can be used as required.
991 delon.nicolas@wanadoo.fr
996 /proc/sys/net/ipv6/* Variables:
998 IPv6 has no global variables such as tcp_*. tcp_* settings under ipv4/ also
999 apply to IPv6 [XXX?].
1001 bindv6only - BOOLEAN
1002 Default value for IPV6_V6ONLY socket option,
1003 which restricts use of the IPv6 socket to IPv6 communication
1005 TRUE: disable IPv4-mapped address feature
1006 FALSE: enable IPv4-mapped address feature
1008 Default: FALSE (as specified in RFC3493)
1012 ip6frag_high_thresh - INTEGER
1013 Maximum memory used to reassemble IPv6 fragments. When
1014 ip6frag_high_thresh bytes of memory is allocated for this purpose,
1015 the fragment handler will toss packets until ip6frag_low_thresh
1018 ip6frag_low_thresh - INTEGER
1019 See ip6frag_high_thresh
1021 ip6frag_time - INTEGER
1022 Time in seconds to keep an IPv6 fragment in memory.
1024 ip6frag_secret_interval - INTEGER
1025 Regeneration interval (in seconds) of the hash secret (or lifetime
1026 for the hash secret) for IPv6 fragments.
1030 Change the interface-specific default settings.
1034 Change all the interface-specific settings.
1036 [XXX: Other special features than forwarding?]
1038 conf/all/forwarding - BOOLEAN
1039 Enable global IPv6 forwarding between all interfaces.
1041 IPv4 and IPv6 work differently here; e.g. netfilter must be used
1042 to control which interfaces may forward packets and which not.
1044 This also sets all interfaces' Host/Router setting
1045 'forwarding' to the specified value. See below for details.
1047 This referred to as global forwarding.
1053 Change special settings per interface.
1055 The functional behaviour for certain settings is different
1056 depending on whether local forwarding is enabled or not.
1059 Accept Router Advertisements; autoconfigure using them.
1061 It also determines whether or not to transmit Router
1062 Solicitations. If and only if the functional setting is to
1063 accept Router Advertisements, Router Solicitations will be
1066 Possible values are:
1067 0 Do not accept Router Advertisements.
1068 1 Accept Router Advertisements if forwarding is disabled.
1069 2 Overrule forwarding behaviour. Accept Router Advertisements
1070 even if forwarding is enabled.
1072 Functional default: enabled if local forwarding is disabled.
1073 disabled if local forwarding is enabled.
1075 accept_ra_defrtr - BOOLEAN
1076 Learn default router in Router Advertisement.
1078 Functional default: enabled if accept_ra is enabled.
1079 disabled if accept_ra is disabled.
1081 accept_ra_min_hop_limit - INTEGER
1082 Minimum hop limit Information in Router Advertisement.
1084 Hop limit Information in Router Advertisement less than this
1085 variable shall be ignored.
1089 accept_ra_pinfo - BOOLEAN
1090 Learn Prefix Information in Router Advertisement.
1092 Functional default: enabled if accept_ra is enabled.
1093 disabled if accept_ra is disabled.
1095 accept_ra_rt_info_max_plen - INTEGER
1096 Maximum prefix length of Route Information in RA.
1098 Route Information w/ prefix larger than or equal to this
1099 variable shall be ignored.
1101 Functional default: 0 if accept_ra_rtr_pref is enabled.
1102 -1 if accept_ra_rtr_pref is disabled.
1104 accept_ra_rtr_pref - BOOLEAN
1105 Accept Router Preference in RA.
1107 Functional default: enabled if accept_ra is enabled.
1108 disabled if accept_ra is disabled.
1110 accept_redirects - BOOLEAN
1113 Functional default: enabled if local forwarding is disabled.
1114 disabled if local forwarding is enabled.
1116 accept_source_route - INTEGER
1117 Accept source routing (routing extension header).
1119 >= 0: Accept only routing header type 2.
1120 < 0: Do not accept routing header.
1125 Autoconfigure addresses using Prefix Information in Router
1128 Functional default: enabled if accept_ra_pinfo is enabled.
1129 disabled if accept_ra_pinfo is disabled.
1131 dad_transmits - INTEGER
1132 The amount of Duplicate Address Detection probes to send.
1135 forwarding - INTEGER
1136 Configure interface-specific Host/Router behaviour.
1138 Note: It is recommended to have the same setting on all
1139 interfaces; mixed router/host scenarios are rather uncommon.
1141 Possible values are:
1142 0 Forwarding disabled
1143 1 Forwarding enabled
1147 By default, Host behaviour is assumed. This means:
1149 1. IsRouter flag is not set in Neighbour Advertisements.
1150 2. If accept_ra is TRUE (default), transmit Router
1152 3. If accept_ra is TRUE (default), accept Router
1153 Advertisements (and do autoconfiguration).
1154 4. If accept_redirects is TRUE (default), accept Redirects.
1158 If local forwarding is enabled, Router behaviour is assumed.
1159 This means exactly the reverse from the above:
1161 1. IsRouter flag is set in Neighbour Advertisements.
1162 2. Router Solicitations are not sent unless accept_ra is 2.
1163 3. Router Advertisements are ignored unless accept_ra is 2.
1164 4. Redirects are ignored.
1166 Default: 0 (disabled) if global forwarding is disabled (default),
1167 otherwise 1 (enabled).
1170 Default Hop Limit to set.
1174 Default Maximum Transfer Unit
1175 Default: 1280 (IPv6 required minimum)
1177 router_probe_interval - INTEGER
1178 Minimum interval (in seconds) between Router Probing described
1183 router_solicitation_delay - INTEGER
1184 Number of seconds to wait after interface is brought up
1185 before sending Router Solicitations.
1188 router_solicitation_interval - INTEGER
1189 Number of seconds to wait between Router Solicitations.
1192 router_solicitations - INTEGER
1193 Number of Router Solicitations to send until assuming no
1194 routers are present.
1197 use_tempaddr - INTEGER
1198 Preference for Privacy Extensions (RFC3041).
1199 <= 0 : disable Privacy Extensions
1200 == 1 : enable Privacy Extensions, but prefer public
1201 addresses over temporary addresses.
1202 > 1 : enable Privacy Extensions and prefer temporary
1203 addresses over public addresses.
1204 Default: 0 (for most devices)
1205 -1 (for point-to-point devices and loopback devices)
1207 temp_valid_lft - INTEGER
1208 valid lifetime (in seconds) for temporary addresses.
1209 Default: 604800 (7 days)
1211 temp_prefered_lft - INTEGER
1212 Preferred lifetime (in seconds) for temporary addresses.
1213 Default: 86400 (1 day)
1215 max_desync_factor - INTEGER
1216 Maximum value for DESYNC_FACTOR, which is a random value
1217 that ensures that clients don't synchronize with each
1218 other and generate new addresses at exactly the same time.
1219 value is in seconds.
1222 regen_max_retry - INTEGER
1223 Number of attempts before give up attempting to generate
1224 valid temporary addresses.
1227 max_addresses - INTEGER
1228 Maximum number of autoconfigured addresses per interface. Setting
1229 to zero disables the limitation. It is not recommended to set this
1230 value too large (or to zero) because it would be an easy way to
1231 crash the kernel by allowing too many addresses to be created.
1234 disable_ipv6 - BOOLEAN
1235 Disable IPv6 operation. If accept_dad is set to 2, this value
1236 will be dynamically set to TRUE if DAD fails for the link-local
1238 Default: FALSE (enable IPv6 operation)
1240 When this value is changed from 1 to 0 (IPv6 is being enabled),
1241 it will dynamically create a link-local address on the given
1242 interface and start Duplicate Address Detection, if necessary.
1244 When this value is changed from 0 to 1 (IPv6 is being disabled),
1245 it will dynamically delete all address on the given interface.
1247 accept_dad - INTEGER
1248 Whether to accept DAD (Duplicate Address Detection).
1250 1: Enable DAD (default)
1251 2: Enable DAD, and disable IPv6 operation if MAC-based duplicate
1252 link-local address has been found.
1254 force_tllao - BOOLEAN
1255 Enable sending the target link-layer address option even when
1256 responding to a unicast neighbor solicitation.
1259 Quoting from RFC 2461, section 4.4, Target link-layer address:
1261 "The option MUST be included for multicast solicitations in order to
1262 avoid infinite Neighbor Solicitation "recursion" when the peer node
1263 does not have a cache entry to return a Neighbor Advertisements
1264 message. When responding to unicast solicitations, the option can be
1265 omitted since the sender of the solicitation has the correct link-
1266 layer address; otherwise it would not have be able to send the unicast
1267 solicitation in the first place. However, including the link-layer
1268 address in this case adds little overhead and eliminates a potential
1269 race condition where the sender deletes the cached link-layer address
1270 prior to receiving a response to a previous solicitation."
1274 Limit the maximal rates for sending ICMPv6 packets.
1275 0 to disable any limiting,
1276 otherwise the minimal space between responses in milliseconds.
1281 Pekka Savola <pekkas@netcore.fi>
1282 YOSHIFUJI Hideaki / USAGI Project <yoshfuji@linux-ipv6.org>
1285 /proc/sys/net/bridge/* Variables:
1287 bridge-nf-call-arptables - BOOLEAN
1288 1 : pass bridged ARP traffic to arptables' FORWARD chain.
1292 bridge-nf-call-iptables - BOOLEAN
1293 1 : pass bridged IPv4 traffic to iptables' chains.
1297 bridge-nf-call-ip6tables - BOOLEAN
1298 1 : pass bridged IPv6 traffic to ip6tables' chains.
1302 bridge-nf-filter-vlan-tagged - BOOLEAN
1303 1 : pass bridged vlan-tagged ARP/IP/IPv6 traffic to {arp,ip,ip6}tables.
1307 bridge-nf-filter-pppoe-tagged - BOOLEAN
1308 1 : pass bridged pppoe-tagged IP/IPv6 traffic to {ip,ip6}tables.
1313 proc/sys/net/sctp/* Variables:
1315 addip_enable - BOOLEAN
1316 Enable or disable extension of Dynamic Address Reconfiguration
1317 (ADD-IP) functionality specified in RFC5061. This extension provides
1318 the ability to dynamically add and remove new addresses for the SCTP
1321 1: Enable extension.
1323 0: Disable extension.
1327 addip_noauth_enable - BOOLEAN
1328 Dynamic Address Reconfiguration (ADD-IP) requires the use of
1329 authentication to protect the operations of adding or removing new
1330 addresses. This requirement is mandated so that unauthorized hosts
1331 would not be able to hijack associations. However, older
1332 implementations may not have implemented this requirement while
1333 allowing the ADD-IP extension. For reasons of interoperability,
1334 we provide this variable to control the enforcement of the
1335 authentication requirement.
1337 1: Allow ADD-IP extension to be used without authentication. This
1338 should only be set in a closed environment for interoperability
1339 with older implementations.
1341 0: Enforce the authentication requirement
1345 auth_enable - BOOLEAN
1346 Enable or disable Authenticated Chunks extension. This extension
1347 provides the ability to send and receive authenticated chunks and is
1348 required for secure operation of Dynamic Address Reconfiguration
1351 1: Enable this extension.
1352 0: Disable this extension.
1356 prsctp_enable - BOOLEAN
1357 Enable or disable the Partial Reliability extension (RFC3758) which
1358 is used to notify peers that a given DATA should no longer be expected.
1366 The limit of the number of new packets that can be initially sent. It
1367 controls how bursty the generated traffic can be.
1371 association_max_retrans - INTEGER
1372 Set the maximum number for retransmissions that an association can
1373 attempt deciding that the remote end is unreachable. If this value
1374 is exceeded, the association is terminated.
1378 max_init_retransmits - INTEGER
1379 The maximum number of retransmissions of INIT and COOKIE-ECHO chunks
1380 that an association will attempt before declaring the destination
1381 unreachable and terminating.
1385 path_max_retrans - INTEGER
1386 The maximum number of retransmissions that will be attempted on a given
1387 path. Once this threshold is exceeded, the path is considered
1388 unreachable, and new traffic will use a different path when the
1389 association is multihomed.
1393 rto_initial - INTEGER
1394 The initial round trip timeout value in milliseconds that will be used
1395 in calculating round trip times. This is the initial time interval
1396 for retransmissions.
1401 The maximum value (in milliseconds) of the round trip timeout. This
1402 is the largest time interval that can elapse between retransmissions.
1407 The minimum value (in milliseconds) of the round trip timeout. This
1408 is the smallest time interval the can elapse between retransmissions.
1412 hb_interval - INTEGER
1413 The interval (in milliseconds) between HEARTBEAT chunks. These chunks
1414 are sent at the specified interval on idle paths to probe the state of
1415 a given path between 2 associations.
1419 sack_timeout - INTEGER
1420 The amount of time (in milliseconds) that the implementation will wait
1425 valid_cookie_life - INTEGER
1426 The default lifetime of the SCTP cookie (in milliseconds). The cookie
1427 is used during association establishment.
1431 cookie_preserve_enable - BOOLEAN
1432 Enable or disable the ability to extend the lifetime of the SCTP cookie
1433 that is used during the establishment phase of SCTP association
1435 1: Enable cookie lifetime extension.
1440 rcvbuf_policy - INTEGER
1441 Determines if the receive buffer is attributed to the socket or to
1442 association. SCTP supports the capability to create multiple
1443 associations on a single socket. When using this capability, it is
1444 possible that a single stalled association that's buffering a lot
1445 of data may block other associations from delivering their data by
1446 consuming all of the receive buffer space. To work around this,
1447 the rcvbuf_policy could be set to attribute the receiver buffer space
1448 to each association instead of the socket. This prevents the described
1451 1: rcvbuf space is per association
1452 0: recbuf space is per socket
1456 sndbuf_policy - INTEGER
1457 Similar to rcvbuf_policy above, this applies to send buffer space.
1459 1: Send buffer is tracked per association
1460 0: Send buffer is tracked per socket.
1464 sctp_mem - vector of 3 INTEGERs: min, pressure, max
1465 Number of pages allowed for queueing by all SCTP sockets.
1467 min: Below this number of pages SCTP is not bothered about its
1468 memory appetite. When amount of memory allocated by SCTP exceeds
1469 this number, SCTP starts to moderate memory usage.
1471 pressure: This value was introduced to follow format of tcp_mem.
1473 max: Number of pages allowed for queueing by all SCTP sockets.
1475 Default is calculated at boot time from amount of available memory.
1477 sctp_rmem - vector of 3 INTEGERs: min, default, max
1478 Only the first value ("min") is used, "default" and "max" are
1481 min: Minimal size of receive buffer used by SCTP socket.
1482 It is guaranteed to each SCTP socket (but not association) even
1483 under moderate memory pressure.
1487 sctp_wmem - vector of 3 INTEGERs: min, default, max
1488 Currently this tunable has no effect.
1490 addr_scope_policy - INTEGER
1491 Control IPv4 address scoping - draft-stewart-tsvwg-sctp-ipv4-00
1493 0 - Disable IPv4 address scoping
1494 1 - Enable IPv4 address scoping
1495 2 - Follow draft but allow IPv4 private addresses
1496 3 - Follow draft but allow IPv4 link local addresses
1501 /proc/sys/net/core/*
1502 dev_weight - INTEGER
1503 The maximum number of packets that kernel can handle on a NAPI
1504 interrupt, it's a Per-CPU variable.
1508 /proc/sys/net/unix/*
1509 max_dgram_qlen - INTEGER
1510 The maximum length of dgram socket receive queue
1517 /proc/sys/net/irda/*
1518 fast_poll_increase FIXME
1519 warn_noreply_time FIXME
1520 discovery_slots FIXME
1523 discovery_timeout FIXME
1524 lap_keepalive_time FIXME
1525 max_noreply_time FIXME
1526 max_tx_data_size FIXME
1528 min_tx_turn_time FIXME