]>
Commit | Line | Data |
---|---|---|
1da177e4 LT |
1 | /proc/sys/net/ipv4/* Variables: |
2 | ||
3 | ip_forward - BOOLEAN | |
4 | 0 - disabled (default) | |
5 | not 0 - enabled | |
6 | ||
7 | Forward Packets between interfaces. | |
8 | ||
9 | This variable is special, its change resets all configuration | |
10 | parameters to their default state (RFC1122 for hosts, RFC1812 | |
11 | for routers) | |
12 | ||
13 | ip_default_ttl - INTEGER | |
14 | default 64 | |
15 | ||
16 | ip_no_pmtu_disc - BOOLEAN | |
17 | Disable Path MTU Discovery. | |
18 | default FALSE | |
19 | ||
20 | min_pmtu - INTEGER | |
21 | default 562 - minimum discovered Path MTU | |
22 | ||
23 | mtu_expires - INTEGER | |
24 | Time, in seconds, that cached PMTU information is kept. | |
25 | ||
26 | min_adv_mss - INTEGER | |
27 | The advertised MSS depends on the first hop route MTU, but will | |
28 | never be lower than this setting. | |
29 | ||
30 | IP Fragmentation: | |
31 | ||
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 | |
36 | is reached. | |
37 | ||
38 | ipfrag_low_thresh - INTEGER | |
39 | See ipfrag_high_thresh | |
40 | ||
41 | ipfrag_time - INTEGER | |
42 | Time in seconds to keep an IP fragment in memory. | |
43 | ||
44 | ipfrag_secret_interval - INTEGER | |
45 | Regeneration interval (in seconds) of the hash secret (or lifetime | |
46 | for the hash secret) for IP fragments. | |
47 | Default: 600 | |
48 | ||
89cee8b1 HX |
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. | |
63 | ||
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. | |
70 | Default: 64 | |
71 | ||
1da177e4 LT |
72 | INET peer storage: |
73 | ||
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. | |
79 | ||
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). | |
85 | ||
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). | |
91 | ||
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). | |
96 | ||
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). | |
101 | ||
102 | TCP variables: | |
103 | ||
ef56e622 SH |
104 | somaxconn - INTEGER |
105 | Limit of socket listen() backlog, known in userspace as SOMAXCONN. | |
106 | Defaults to 128. See also tcp_max_syn_backlog for additional tuning | |
107 | for TCP sockets. | |
108 | ||
9772efb9 | 109 | tcp_abc - INTEGER |
b3a8a40d SH |
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. | |
113 | Possible values are: | |
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. | |
118 | Default: 0 (off) | |
9772efb9 | 119 | |
ef56e622 SH |
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. | |
1da177e4 | 127 | |
ef56e622 SH |
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), | |
131 | if it is <= 0. | |
132 | Default: 2 | |
1da177e4 | 133 | |
ef56e622 SH |
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). | |
1da177e4 | 139 | |
ef56e622 SH |
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. | |
143 | Default: 31 | |
1da177e4 | 144 | |
ef56e622 SH |
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, | |
148 | but not loaded. | |
1da177e4 | 149 | |
71599cd1 JH |
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. | |
154 | ||
ef56e622 SH |
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. | |
1da177e4 | 160 | |
ef56e622 SH |
161 | tcp_dsack - BOOLEAN |
162 | Allows TCP to send "duplicate" SACKs. | |
1da177e4 | 163 | |
ef56e622 SH |
164 | tcp_ecn - BOOLEAN |
165 | Enable Explicit Congestion Notification in TCP. | |
166 | ||
167 | tcp_fack - BOOLEAN | |
168 | Enable FACK congestion avoidance and fast retransmission. | |
169 | The value is not used, if tcp_sack is not enabled. | |
1da177e4 LT |
170 | |
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. | |
181 | ||
ef56e622 SH |
182 | tcp_frto - BOOLEAN |
183 | Enables F-RTO, an enhanced recovery algorithm for TCP retransmission | |
184 | timeouts. It is particularly beneficial in wireless environments | |
185 | where packet loss is typically due to random radio interference | |
127af0c4 IJ |
186 | rather than intermediate router congestion. If set to 1, basic |
187 | version is enabled. 2 enables SACK enhanced FRTO, which is | |
188 | EXPERIMENTAL. The basic version can be used also when SACK is | |
189 | enabled for a flow through tcp_sack sysctl. | |
1da177e4 | 190 | |
ef56e622 SH |
191 | tcp_keepalive_time - INTEGER |
192 | How often TCP sends out keepalive messages when keepalive is enabled. | |
193 | Default: 2hours. | |
1da177e4 | 194 | |
ef56e622 SH |
195 | tcp_keepalive_probes - INTEGER |
196 | How many keepalive probes TCP sends out, until it decides that the | |
197 | connection is broken. Default value: 9. | |
198 | ||
199 | tcp_keepalive_intvl - INTEGER | |
200 | How frequently the probes are send out. Multiplied by | |
201 | tcp_keepalive_probes it is time to kill not responding connection, | |
202 | after probes started. Default value: 75sec i.e. connection | |
203 | will be aborted after ~11 minutes of retries. | |
204 | ||
205 | tcp_low_latency - BOOLEAN | |
206 | If set, the TCP stack makes decisions that prefer lower | |
207 | latency as opposed to higher throughput. By default, this | |
208 | option is not set meaning that higher throughput is preferred. | |
209 | An example of an application where this default should be | |
210 | changed would be a Beowulf compute cluster. | |
211 | Default: 0 | |
1da177e4 LT |
212 | |
213 | tcp_max_orphans - INTEGER | |
214 | Maximal number of TCP sockets not attached to any user file handle, | |
215 | held by system. If this number is exceeded orphaned connections are | |
216 | reset immediately and warning is printed. This limit exists | |
217 | only to prevent simple DoS attacks, you _must_ not rely on this | |
218 | or lower the limit artificially, but rather increase it | |
219 | (probably, after increasing installed memory), | |
220 | if network conditions require more than default value, | |
221 | and tune network services to linger and kill such states | |
222 | more aggressively. Let me to remind again: each orphan eats | |
223 | up to ~64K of unswappable memory. | |
224 | ||
1da177e4 LT |
225 | tcp_max_syn_backlog - INTEGER |
226 | Maximal number of remembered connection requests, which are | |
227 | still did not receive an acknowledgment from connecting client. | |
228 | Default value is 1024 for systems with more than 128Mb of memory, | |
229 | and 128 for low memory machines. If server suffers of overload, | |
230 | try to increase this number. | |
231 | ||
ef56e622 SH |
232 | tcp_max_tw_buckets - INTEGER |
233 | Maximal number of timewait sockets held by system simultaneously. | |
234 | If this number is exceeded time-wait socket is immediately destroyed | |
235 | and warning is printed. This limit exists only to prevent | |
236 | simple DoS attacks, you _must_ not lower the limit artificially, | |
237 | but rather increase it (probably, after increasing installed memory), | |
238 | if network conditions require more than default value. | |
1da177e4 | 239 | |
ef56e622 SH |
240 | tcp_mem - vector of 3 INTEGERs: min, pressure, max |
241 | min: below this number of pages TCP is not bothered about its | |
242 | memory appetite. | |
1da177e4 | 243 | |
ef56e622 SH |
244 | pressure: when amount of memory allocated by TCP exceeds this number |
245 | of pages, TCP moderates its memory consumption and enters memory | |
246 | pressure mode, which is exited when memory consumption falls | |
247 | under "min". | |
1da177e4 | 248 | |
ef56e622 | 249 | max: number of pages allowed for queueing by all TCP sockets. |
1da177e4 | 250 | |
ef56e622 SH |
251 | Defaults are calculated at boot time from amount of available |
252 | memory. | |
1da177e4 | 253 | |
71599cd1 JH |
254 | tcp_moderate_rcvbuf - BOOLEAN |
255 | If set, TCP performs receive buffer autotuning, attempting to | |
256 | automatically size the buffer (no greater than tcp_rmem[2]) to | |
257 | match the size required by the path for full throughput. Enabled by | |
258 | default. | |
259 | ||
260 | tcp_mtu_probing - INTEGER | |
261 | Controls TCP Packetization-Layer Path MTU Discovery. Takes three | |
262 | values: | |
263 | 0 - Disabled | |
264 | 1 - Disabled by default, enabled when an ICMP black hole detected | |
265 | 2 - Always enabled, use initial MSS of tcp_base_mss. | |
266 | ||
267 | tcp_no_metrics_save - BOOLEAN | |
268 | By default, TCP saves various connection metrics in the route cache | |
269 | when the connection closes, so that connections established in the | |
270 | near future can use these to set initial conditions. Usually, this | |
271 | increases overall performance, but may sometimes cause performance | |
272 | degredation. If set, TCP will not cache metrics on closing | |
273 | connections. | |
274 | ||
ef56e622 SH |
275 | tcp_orphan_retries - INTEGER |
276 | How may times to retry before killing TCP connection, closed | |
277 | by our side. Default value 7 corresponds to ~50sec-16min | |
278 | depending on RTO. If you machine is loaded WEB server, | |
279 | you should think about lowering this value, such sockets | |
280 | may consume significant resources. Cf. tcp_max_orphans. | |
1da177e4 LT |
281 | |
282 | tcp_reordering - INTEGER | |
283 | Maximal reordering of packets in a TCP stream. | |
284 | Default: 3 | |
285 | ||
286 | tcp_retrans_collapse - BOOLEAN | |
287 | Bug-to-bug compatibility with some broken printers. | |
288 | On retransmit try to send bigger packets to work around bugs in | |
289 | certain TCP stacks. | |
290 | ||
ef56e622 SH |
291 | tcp_retries1 - INTEGER |
292 | How many times to retry before deciding that something is wrong | |
293 | and it is necessary to report this suspicion to network layer. | |
294 | Minimal RFC value is 3, it is default, which corresponds | |
295 | to ~3sec-8min depending on RTO. | |
1da177e4 | 296 | |
ef56e622 SH |
297 | tcp_retries2 - INTEGER |
298 | How may times to retry before killing alive TCP connection. | |
299 | RFC1122 says that the limit should be longer than 100 sec. | |
300 | It is too small number. Default value 15 corresponds to ~13-30min | |
301 | depending on RTO. | |
1da177e4 | 302 | |
ef56e622 SH |
303 | tcp_rfc1337 - BOOLEAN |
304 | If set, the TCP stack behaves conforming to RFC1337. If unset, | |
305 | we are not conforming to RFC, but prevent TCP TIME_WAIT | |
306 | assassination. | |
307 | Default: 0 | |
1da177e4 LT |
308 | |
309 | tcp_rmem - vector of 3 INTEGERs: min, default, max | |
310 | min: Minimal size of receive buffer used by TCP sockets. | |
311 | It is guaranteed to each TCP socket, even under moderate memory | |
312 | pressure. | |
313 | Default: 8K | |
314 | ||
315 | default: default size of receive buffer used by TCP sockets. | |
316 | This value overrides net.core.rmem_default used by other protocols. | |
317 | Default: 87380 bytes. This value results in window of 65535 with | |
318 | default setting of tcp_adv_win_scale and tcp_app_win:0 and a bit | |
319 | less for default tcp_app_win. See below about these variables. | |
320 | ||
321 | max: maximal size of receive buffer allowed for automatically | |
322 | selected receiver buffers for TCP socket. This value does not override | |
323 | net.core.rmem_max, "static" selection via SO_RCVBUF does not use this. | |
324 | Default: 87380*2 bytes. | |
325 | ||
ef56e622 SH |
326 | tcp_sack - BOOLEAN |
327 | Enable select acknowledgments (SACKS). | |
1da177e4 | 328 | |
ef56e622 SH |
329 | tcp_slow_start_after_idle - BOOLEAN |
330 | If set, provide RFC2861 behavior and time out the congestion | |
331 | window after an idle period. An idle period is defined at | |
332 | the current RTO. If unset, the congestion window will not | |
333 | be timed out after an idle period. | |
334 | Default: 1 | |
1da177e4 | 335 | |
ef56e622 SH |
336 | tcp_stdurg - BOOLEAN |
337 | Use the Host requirements interpretation of the TCP urg pointer field. | |
338 | Most hosts use the older BSD interpretation, so if you turn this on | |
339 | Linux might not communicate correctly with them. | |
340 | Default: FALSE | |
1da177e4 | 341 | |
ef56e622 SH |
342 | tcp_synack_retries - INTEGER |
343 | Number of times SYNACKs for a passive TCP connection attempt will | |
344 | be retransmitted. Should not be higher than 255. Default value | |
345 | is 5, which corresponds to ~180seconds. | |
1da177e4 | 346 | |
ef56e622 SH |
347 | tcp_syncookies - BOOLEAN |
348 | Only valid when the kernel was compiled with CONFIG_SYNCOOKIES | |
349 | Send out syncookies when the syn backlog queue of a socket | |
350 | overflows. This is to prevent against the common 'syn flood attack' | |
351 | Default: FALSE | |
1da177e4 | 352 | |
ef56e622 SH |
353 | Note, that syncookies is fallback facility. |
354 | It MUST NOT be used to help highly loaded servers to stand | |
355 | against legal connection rate. If you see synflood warnings | |
356 | in your logs, but investigation shows that they occur | |
357 | because of overload with legal connections, you should tune | |
358 | another parameters until this warning disappear. | |
359 | See: tcp_max_syn_backlog, tcp_synack_retries, tcp_abort_on_overflow. | |
1da177e4 | 360 | |
ef56e622 SH |
361 | syncookies seriously violate TCP protocol, do not allow |
362 | to use TCP extensions, can result in serious degradation | |
363 | of some services (f.e. SMTP relaying), visible not by you, | |
364 | but your clients and relays, contacting you. While you see | |
365 | synflood warnings in logs not being really flooded, your server | |
366 | is seriously misconfigured. | |
1da177e4 | 367 | |
ef56e622 SH |
368 | tcp_syn_retries - INTEGER |
369 | Number of times initial SYNs for an active TCP connection attempt | |
370 | will be retransmitted. Should not be higher than 255. Default value | |
371 | is 5, which corresponds to ~180seconds. | |
372 | ||
373 | tcp_timestamps - BOOLEAN | |
374 | Enable timestamps as defined in RFC1323. | |
1da177e4 | 375 | |
1da177e4 | 376 | tcp_tso_win_divisor - INTEGER |
ef56e622 SH |
377 | This allows control over what percentage of the congestion window |
378 | can be consumed by a single TSO frame. | |
379 | The setting of this parameter is a choice between burstiness and | |
380 | building larger TSO frames. | |
381 | Default: 3 | |
1da177e4 | 382 | |
ef56e622 SH |
383 | tcp_tw_recycle - BOOLEAN |
384 | Enable fast recycling TIME-WAIT sockets. Default value is 0. | |
385 | It should not be changed without advice/request of technical | |
386 | experts. | |
1da177e4 | 387 | |
ef56e622 SH |
388 | tcp_tw_reuse - BOOLEAN |
389 | Allow to reuse TIME-WAIT sockets for new connections when it is | |
390 | safe from protocol viewpoint. Default value is 0. | |
391 | It should not be changed without advice/request of technical | |
392 | experts. | |
ce7bc3bf | 393 | |
ef56e622 SH |
394 | tcp_window_scaling - BOOLEAN |
395 | Enable window scaling as defined in RFC1323. | |
3ff825b2 | 396 | |
ef56e622 SH |
397 | tcp_wmem - vector of 3 INTEGERs: min, default, max |
398 | min: Amount of memory reserved for send buffers for TCP socket. | |
399 | Each TCP socket has rights to use it due to fact of its birth. | |
400 | Default: 4K | |
9d7bcfc6 | 401 | |
ef56e622 SH |
402 | default: Amount of memory allowed for send buffers for TCP socket |
403 | by default. This value overrides net.core.wmem_default used | |
404 | by other protocols, it is usually lower than net.core.wmem_default. | |
405 | Default: 16K | |
406 | ||
407 | max: Maximal amount of memory allowed for automatically selected | |
408 | send buffers for TCP socket. This value does not override | |
409 | net.core.wmem_max, "static" selection via SO_SNDBUF does not use this. | |
410 | Default: 128K | |
1da177e4 | 411 | |
15d99e02 RJ |
412 | tcp_workaround_signed_windows - BOOLEAN |
413 | If set, assume no receipt of a window scaling option means the | |
414 | remote TCP is broken and treats the window as a signed quantity. | |
415 | If unset, assume the remote TCP is not broken even if we do | |
416 | not receive a window scaling option from them. | |
417 | Default: 0 | |
418 | ||
8802f616 PM |
419 | CIPSOv4 Variables: |
420 | ||
421 | cipso_cache_enable - BOOLEAN | |
422 | If set, enable additions to and lookups from the CIPSO label mapping | |
423 | cache. If unset, additions are ignored and lookups always result in a | |
424 | miss. However, regardless of the setting the cache is still | |
425 | invalidated when required when means you can safely toggle this on and | |
426 | off and the cache will always be "safe". | |
427 | Default: 1 | |
428 | ||
429 | cipso_cache_bucket_size - INTEGER | |
430 | The CIPSO label cache consists of a fixed size hash table with each | |
431 | hash bucket containing a number of cache entries. This variable limits | |
432 | the number of entries in each hash bucket; the larger the value the | |
433 | more CIPSO label mappings that can be cached. When the number of | |
434 | entries in a given hash bucket reaches this limit adding new entries | |
435 | causes the oldest entry in the bucket to be removed to make room. | |
436 | Default: 10 | |
437 | ||
438 | cipso_rbm_optfmt - BOOLEAN | |
439 | Enable the "Optimized Tag 1 Format" as defined in section 3.4.2.6 of | |
440 | the CIPSO draft specification (see Documentation/netlabel for details). | |
441 | This means that when set the CIPSO tag will be padded with empty | |
442 | categories in order to make the packet data 32-bit aligned. | |
443 | Default: 0 | |
444 | ||
445 | cipso_rbm_structvalid - BOOLEAN | |
446 | If set, do a very strict check of the CIPSO option when | |
447 | ip_options_compile() is called. If unset, relax the checks done during | |
448 | ip_options_compile(). Either way is "safe" as errors are caught else | |
449 | where in the CIPSO processing code but setting this to 0 (False) should | |
450 | result in less work (i.e. it should be faster) but could cause problems | |
451 | with other implementations that require strict checking. | |
452 | Default: 0 | |
453 | ||
1da177e4 LT |
454 | IP Variables: |
455 | ||
456 | ip_local_port_range - 2 INTEGERS | |
457 | Defines the local port range that is used by TCP and UDP to | |
458 | choose the local port. The first number is the first, the | |
459 | second the last local port number. Default value depends on | |
460 | amount of memory available on the system: | |
461 | > 128Mb 32768-61000 | |
462 | < 128Mb 1024-4999 or even less. | |
463 | This number defines number of active connections, which this | |
464 | system can issue simultaneously to systems not supporting | |
465 | TCP extensions (timestamps). With tcp_tw_recycle enabled | |
466 | (i.e. by default) range 1024-4999 is enough to issue up to | |
467 | 2000 connections per second to systems supporting timestamps. | |
468 | ||
469 | ip_nonlocal_bind - BOOLEAN | |
470 | If set, allows processes to bind() to non-local IP addresses, | |
471 | which can be quite useful - but may break some applications. | |
472 | Default: 0 | |
473 | ||
474 | ip_dynaddr - BOOLEAN | |
475 | If set non-zero, enables support for dynamic addresses. | |
476 | If set to a non-zero value larger than 1, a kernel log | |
477 | message will be printed when dynamic address rewriting | |
478 | occurs. | |
479 | Default: 0 | |
480 | ||
481 | icmp_echo_ignore_all - BOOLEAN | |
7ce31246 DM |
482 | If set non-zero, then the kernel will ignore all ICMP ECHO |
483 | requests sent to it. | |
484 | Default: 0 | |
485 | ||
1da177e4 | 486 | icmp_echo_ignore_broadcasts - BOOLEAN |
7ce31246 DM |
487 | If set non-zero, then the kernel will ignore all ICMP ECHO and |
488 | TIMESTAMP requests sent to it via broadcast/multicast. | |
489 | Default: 1 | |
1da177e4 LT |
490 | |
491 | icmp_ratelimit - INTEGER | |
492 | Limit the maximal rates for sending ICMP packets whose type matches | |
493 | icmp_ratemask (see below) to specific targets. | |
494 | 0 to disable any limiting, otherwise the maximal rate in jiffies(1) | |
495 | Default: 100 | |
496 | ||
497 | icmp_ratemask - INTEGER | |
498 | Mask made of ICMP types for which rates are being limited. | |
499 | Significant bits: IHGFEDCBA9876543210 | |
500 | Default mask: 0000001100000011000 (6168) | |
501 | ||
502 | Bit definitions (see include/linux/icmp.h): | |
503 | 0 Echo Reply | |
504 | 3 Destination Unreachable * | |
505 | 4 Source Quench * | |
506 | 5 Redirect | |
507 | 8 Echo Request | |
508 | B Time Exceeded * | |
509 | C Parameter Problem * | |
510 | D Timestamp Request | |
511 | E Timestamp Reply | |
512 | F Info Request | |
513 | G Info Reply | |
514 | H Address Mask Request | |
515 | I Address Mask Reply | |
516 | ||
517 | * These are rate limited by default (see default mask above) | |
518 | ||
519 | icmp_ignore_bogus_error_responses - BOOLEAN | |
520 | Some routers violate RFC1122 by sending bogus responses to broadcast | |
521 | frames. Such violations are normally logged via a kernel warning. | |
522 | If this is set to TRUE, the kernel will not give such warnings, which | |
523 | will avoid log file clutter. | |
524 | Default: FALSE | |
525 | ||
95f7daf1 H |
526 | icmp_errors_use_inbound_ifaddr - BOOLEAN |
527 | ||
528 | If zero, icmp error messages are sent with the primary address of | |
529 | the exiting interface. | |
530 | ||
531 | If non-zero, the message will be sent with the primary address of | |
532 | the interface that received the packet that caused the icmp error. | |
533 | This is the behaviour network many administrators will expect from | |
534 | a router. And it can make debugging complicated network layouts | |
535 | much easier. | |
536 | ||
537 | Note that if no primary address exists for the interface selected, | |
538 | then the primary address of the first non-loopback interface that | |
d6bc8ac9 | 539 | has one will be used regardless of this setting. |
95f7daf1 H |
540 | |
541 | Default: 0 | |
542 | ||
1da177e4 LT |
543 | igmp_max_memberships - INTEGER |
544 | Change the maximum number of multicast groups we can subscribe to. | |
545 | Default: 20 | |
546 | ||
547 | conf/interface/* changes special settings per interface (where "interface" is | |
548 | the name of your network interface) | |
549 | conf/all/* is special, changes the settings for all interfaces | |
550 | ||
551 | ||
552 | log_martians - BOOLEAN | |
553 | Log packets with impossible addresses to kernel log. | |
554 | log_martians for the interface will be enabled if at least one of | |
555 | conf/{all,interface}/log_martians is set to TRUE, | |
556 | it will be disabled otherwise | |
557 | ||
558 | accept_redirects - BOOLEAN | |
559 | Accept ICMP redirect messages. | |
560 | accept_redirects for the interface will be enabled if: | |
561 | - both conf/{all,interface}/accept_redirects are TRUE in the case forwarding | |
562 | for the interface is enabled | |
563 | or | |
564 | - at least one of conf/{all,interface}/accept_redirects is TRUE in the case | |
565 | forwarding for the interface is disabled | |
566 | accept_redirects for the interface will be disabled otherwise | |
567 | default TRUE (host) | |
568 | FALSE (router) | |
569 | ||
570 | forwarding - BOOLEAN | |
571 | Enable IP forwarding on this interface. | |
572 | ||
573 | mc_forwarding - BOOLEAN | |
574 | Do multicast routing. The kernel needs to be compiled with CONFIG_MROUTE | |
575 | and a multicast routing daemon is required. | |
576 | conf/all/mc_forwarding must also be set to TRUE to enable multicast routing | |
577 | for the interface | |
578 | ||
579 | medium_id - INTEGER | |
580 | Integer value used to differentiate the devices by the medium they | |
581 | are attached to. Two devices can have different id values when | |
582 | the broadcast packets are received only on one of them. | |
583 | The default value 0 means that the device is the only interface | |
584 | to its medium, value of -1 means that medium is not known. | |
585 | ||
586 | Currently, it is used to change the proxy_arp behavior: | |
587 | the proxy_arp feature is enabled for packets forwarded between | |
588 | two devices attached to different media. | |
589 | ||
590 | proxy_arp - BOOLEAN | |
591 | Do proxy arp. | |
592 | proxy_arp for the interface will be enabled if at least one of | |
593 | conf/{all,interface}/proxy_arp is set to TRUE, | |
594 | it will be disabled otherwise | |
595 | ||
596 | shared_media - BOOLEAN | |
597 | Send(router) or accept(host) RFC1620 shared media redirects. | |
598 | Overrides ip_secure_redirects. | |
599 | shared_media for the interface will be enabled if at least one of | |
600 | conf/{all,interface}/shared_media is set to TRUE, | |
601 | it will be disabled otherwise | |
602 | default TRUE | |
603 | ||
604 | secure_redirects - BOOLEAN | |
605 | Accept ICMP redirect messages only for gateways, | |
606 | listed in default gateway list. | |
607 | secure_redirects for the interface will be enabled if at least one of | |
608 | conf/{all,interface}/secure_redirects is set to TRUE, | |
609 | it will be disabled otherwise | |
610 | default TRUE | |
611 | ||
612 | send_redirects - BOOLEAN | |
613 | Send redirects, if router. | |
614 | send_redirects for the interface will be enabled if at least one of | |
615 | conf/{all,interface}/send_redirects is set to TRUE, | |
616 | it will be disabled otherwise | |
617 | Default: TRUE | |
618 | ||
619 | bootp_relay - BOOLEAN | |
620 | Accept packets with source address 0.b.c.d destined | |
621 | not to this host as local ones. It is supposed, that | |
622 | BOOTP relay daemon will catch and forward such packets. | |
623 | conf/all/bootp_relay must also be set to TRUE to enable BOOTP relay | |
624 | for the interface | |
625 | default FALSE | |
626 | Not Implemented Yet. | |
627 | ||
628 | accept_source_route - BOOLEAN | |
629 | Accept packets with SRR option. | |
630 | conf/all/accept_source_route must also be set to TRUE to accept packets | |
631 | with SRR option on the interface | |
632 | default TRUE (router) | |
633 | FALSE (host) | |
634 | ||
635 | rp_filter - BOOLEAN | |
636 | 1 - do source validation by reversed path, as specified in RFC1812 | |
637 | Recommended option for single homed hosts and stub network | |
638 | routers. Could cause troubles for complicated (not loop free) | |
639 | networks running a slow unreliable protocol (sort of RIP), | |
640 | or using static routes. | |
641 | ||
642 | 0 - No source validation. | |
643 | ||
644 | conf/all/rp_filter must also be set to TRUE to do source validation | |
645 | on the interface | |
646 | ||
647 | Default value is 0. Note that some distributions enable it | |
648 | in startup scripts. | |
649 | ||
650 | arp_filter - BOOLEAN | |
651 | 1 - Allows you to have multiple network interfaces on the same | |
652 | subnet, and have the ARPs for each interface be answered | |
653 | based on whether or not the kernel would route a packet from | |
654 | the ARP'd IP out that interface (therefore you must use source | |
655 | based routing for this to work). In other words it allows control | |
656 | of which cards (usually 1) will respond to an arp request. | |
657 | ||
658 | 0 - (default) The kernel can respond to arp requests with addresses | |
659 | from other interfaces. This may seem wrong but it usually makes | |
660 | sense, because it increases the chance of successful communication. | |
661 | IP addresses are owned by the complete host on Linux, not by | |
662 | particular interfaces. Only for more complex setups like load- | |
663 | balancing, does this behaviour cause problems. | |
664 | ||
665 | arp_filter for the interface will be enabled if at least one of | |
666 | conf/{all,interface}/arp_filter is set to TRUE, | |
667 | it will be disabled otherwise | |
668 | ||
669 | arp_announce - INTEGER | |
670 | Define different restriction levels for announcing the local | |
671 | source IP address from IP packets in ARP requests sent on | |
672 | interface: | |
673 | 0 - (default) Use any local address, configured on any interface | |
674 | 1 - Try to avoid local addresses that are not in the target's | |
675 | subnet for this interface. This mode is useful when target | |
676 | hosts reachable via this interface require the source IP | |
677 | address in ARP requests to be part of their logical network | |
678 | configured on the receiving interface. When we generate the | |
679 | request we will check all our subnets that include the | |
680 | target IP and will preserve the source address if it is from | |
681 | such subnet. If there is no such subnet we select source | |
682 | address according to the rules for level 2. | |
683 | 2 - Always use the best local address for this target. | |
684 | In this mode we ignore the source address in the IP packet | |
685 | and try to select local address that we prefer for talks with | |
686 | the target host. Such local address is selected by looking | |
687 | for primary IP addresses on all our subnets on the outgoing | |
688 | interface that include the target IP address. If no suitable | |
689 | local address is found we select the first local address | |
690 | we have on the outgoing interface or on all other interfaces, | |
691 | with the hope we will receive reply for our request and | |
692 | even sometimes no matter the source IP address we announce. | |
693 | ||
694 | The max value from conf/{all,interface}/arp_announce is used. | |
695 | ||
696 | Increasing the restriction level gives more chance for | |
697 | receiving answer from the resolved target while decreasing | |
698 | the level announces more valid sender's information. | |
699 | ||
700 | arp_ignore - INTEGER | |
701 | Define different modes for sending replies in response to | |
702 | received ARP requests that resolve local target IP addresses: | |
703 | 0 - (default): reply for any local target IP address, configured | |
704 | on any interface | |
705 | 1 - reply only if the target IP address is local address | |
706 | configured on the incoming interface | |
707 | 2 - reply only if the target IP address is local address | |
708 | configured on the incoming interface and both with the | |
709 | sender's IP address are part from same subnet on this interface | |
710 | 3 - do not reply for local addresses configured with scope host, | |
711 | only resolutions for global and link addresses are replied | |
712 | 4-7 - reserved | |
713 | 8 - do not reply for all local addresses | |
714 | ||
715 | The max value from conf/{all,interface}/arp_ignore is used | |
716 | when ARP request is received on the {interface} | |
717 | ||
c1b1bce8 NH |
718 | arp_accept - BOOLEAN |
719 | Define behavior when gratuitous arp replies are received: | |
720 | 0 - drop gratuitous arp frames | |
721 | 1 - accept gratuitous arp frames | |
722 | ||
1da177e4 LT |
723 | app_solicit - INTEGER |
724 | The maximum number of probes to send to the user space ARP daemon | |
725 | via netlink before dropping back to multicast probes (see | |
726 | mcast_solicit). Defaults to 0. | |
727 | ||
728 | disable_policy - BOOLEAN | |
729 | Disable IPSEC policy (SPD) for this interface | |
730 | ||
731 | disable_xfrm - BOOLEAN | |
732 | Disable IPSEC encryption on this interface, whatever the policy | |
733 | ||
734 | ||
735 | ||
736 | tag - INTEGER | |
737 | Allows you to write a number, which can be used as required. | |
738 | Default value is 0. | |
739 | ||
740 | (1) Jiffie: internal timeunit for the kernel. On the i386 1/100s, on the | |
741 | Alpha 1/1024s. See the HZ define in /usr/include/asm/param.h for the exact | |
742 | value on your system. | |
743 | ||
744 | Alexey Kuznetsov. | |
745 | kuznet@ms2.inr.ac.ru | |
746 | ||
747 | Updated by: | |
748 | Andi Kleen | |
749 | ak@muc.de | |
750 | Nicolas Delon | |
751 | delon.nicolas@wanadoo.fr | |
752 | ||
753 | ||
754 | ||
755 | ||
756 | /proc/sys/net/ipv6/* Variables: | |
757 | ||
758 | IPv6 has no global variables such as tcp_*. tcp_* settings under ipv4/ also | |
759 | apply to IPv6 [XXX?]. | |
760 | ||
761 | bindv6only - BOOLEAN | |
762 | Default value for IPV6_V6ONLY socket option, | |
763 | which restricts use of the IPv6 socket to IPv6 communication | |
764 | only. | |
765 | TRUE: disable IPv4-mapped address feature | |
766 | FALSE: enable IPv4-mapped address feature | |
767 | ||
768 | Default: FALSE (as specified in RFC2553bis) | |
769 | ||
770 | IPv6 Fragmentation: | |
771 | ||
772 | ip6frag_high_thresh - INTEGER | |
773 | Maximum memory used to reassemble IPv6 fragments. When | |
774 | ip6frag_high_thresh bytes of memory is allocated for this purpose, | |
775 | the fragment handler will toss packets until ip6frag_low_thresh | |
776 | is reached. | |
777 | ||
778 | ip6frag_low_thresh - INTEGER | |
779 | See ip6frag_high_thresh | |
780 | ||
781 | ip6frag_time - INTEGER | |
782 | Time in seconds to keep an IPv6 fragment in memory. | |
783 | ||
784 | ip6frag_secret_interval - INTEGER | |
785 | Regeneration interval (in seconds) of the hash secret (or lifetime | |
786 | for the hash secret) for IPv6 fragments. | |
787 | Default: 600 | |
788 | ||
789 | conf/default/*: | |
790 | Change the interface-specific default settings. | |
791 | ||
792 | ||
793 | conf/all/*: | |
794 | Change all the interface-specific settings. | |
795 | ||
796 | [XXX: Other special features than forwarding?] | |
797 | ||
798 | conf/all/forwarding - BOOLEAN | |
799 | Enable global IPv6 forwarding between all interfaces. | |
800 | ||
801 | IPv4 and IPv6 work differently here; e.g. netfilter must be used | |
802 | to control which interfaces may forward packets and which not. | |
803 | ||
804 | This also sets all interfaces' Host/Router setting | |
805 | 'forwarding' to the specified value. See below for details. | |
806 | ||
807 | This referred to as global forwarding. | |
808 | ||
fbea49e1 YH |
809 | proxy_ndp - BOOLEAN |
810 | Do proxy ndp. | |
811 | ||
1da177e4 LT |
812 | conf/interface/*: |
813 | Change special settings per interface. | |
814 | ||
815 | The functional behaviour for certain settings is different | |
816 | depending on whether local forwarding is enabled or not. | |
817 | ||
818 | accept_ra - BOOLEAN | |
819 | Accept Router Advertisements; autoconfigure using them. | |
820 | ||
821 | Functional default: enabled if local forwarding is disabled. | |
822 | disabled if local forwarding is enabled. | |
823 | ||
65f5c7c1 YH |
824 | accept_ra_defrtr - BOOLEAN |
825 | Learn default router in Router Advertisement. | |
826 | ||
827 | Functional default: enabled if accept_ra is enabled. | |
828 | disabled if accept_ra is disabled. | |
829 | ||
c4fd30eb | 830 | accept_ra_pinfo - BOOLEAN |
2fe0ae78 | 831 | Learn Prefix Information in Router Advertisement. |
c4fd30eb YH |
832 | |
833 | Functional default: enabled if accept_ra is enabled. | |
834 | disabled if accept_ra is disabled. | |
835 | ||
09c884d4 YH |
836 | accept_ra_rt_info_max_plen - INTEGER |
837 | Maximum prefix length of Route Information in RA. | |
838 | ||
839 | Route Information w/ prefix larger than or equal to this | |
840 | variable shall be ignored. | |
841 | ||
842 | Functional default: 0 if accept_ra_rtr_pref is enabled. | |
843 | -1 if accept_ra_rtr_pref is disabled. | |
844 | ||
930d6ff2 YH |
845 | accept_ra_rtr_pref - BOOLEAN |
846 | Accept Router Preference in RA. | |
847 | ||
848 | Functional default: enabled if accept_ra is enabled. | |
849 | disabled if accept_ra is disabled. | |
850 | ||
1da177e4 LT |
851 | accept_redirects - BOOLEAN |
852 | Accept Redirects. | |
853 | ||
854 | Functional default: enabled if local forwarding is disabled. | |
855 | disabled if local forwarding is enabled. | |
856 | ||
0bcbc926 YH |
857 | accept_source_route - INTEGER |
858 | Accept source routing (routing extension header). | |
859 | ||
860 | > 0: Accept routing header. | |
861 | = 0: Accept only routing header type 2. | |
862 | < 0: Do not accept routing header. | |
863 | ||
864 | Default: 0 | |
865 | ||
1da177e4 LT |
866 | autoconf - BOOLEAN |
867 | Autoconfigure addresses using Prefix Information in Router | |
868 | Advertisements. | |
869 | ||
c4fd30eb YH |
870 | Functional default: enabled if accept_ra_pinfo is enabled. |
871 | disabled if accept_ra_pinfo is disabled. | |
1da177e4 LT |
872 | |
873 | dad_transmits - INTEGER | |
874 | The amount of Duplicate Address Detection probes to send. | |
875 | Default: 1 | |
876 | ||
877 | forwarding - BOOLEAN | |
878 | Configure interface-specific Host/Router behaviour. | |
879 | ||
880 | Note: It is recommended to have the same setting on all | |
881 | interfaces; mixed router/host scenarios are rather uncommon. | |
882 | ||
883 | FALSE: | |
884 | ||
885 | By default, Host behaviour is assumed. This means: | |
886 | ||
887 | 1. IsRouter flag is not set in Neighbour Advertisements. | |
888 | 2. Router Solicitations are being sent when necessary. | |
889 | 3. If accept_ra is TRUE (default), accept Router | |
890 | Advertisements (and do autoconfiguration). | |
891 | 4. If accept_redirects is TRUE (default), accept Redirects. | |
892 | ||
893 | TRUE: | |
894 | ||
895 | If local forwarding is enabled, Router behaviour is assumed. | |
896 | This means exactly the reverse from the above: | |
897 | ||
898 | 1. IsRouter flag is set in Neighbour Advertisements. | |
899 | 2. Router Solicitations are not sent. | |
900 | 3. Router Advertisements are ignored. | |
901 | 4. Redirects are ignored. | |
902 | ||
903 | Default: FALSE if global forwarding is disabled (default), | |
904 | otherwise TRUE. | |
905 | ||
906 | hop_limit - INTEGER | |
907 | Default Hop Limit to set. | |
908 | Default: 64 | |
909 | ||
910 | mtu - INTEGER | |
911 | Default Maximum Transfer Unit | |
912 | Default: 1280 (IPv6 required minimum) | |
913 | ||
52e16356 YH |
914 | router_probe_interval - INTEGER |
915 | Minimum interval (in seconds) between Router Probing described | |
916 | in RFC4191. | |
917 | ||
918 | Default: 60 | |
919 | ||
1da177e4 LT |
920 | router_solicitation_delay - INTEGER |
921 | Number of seconds to wait after interface is brought up | |
922 | before sending Router Solicitations. | |
923 | Default: 1 | |
924 | ||
925 | router_solicitation_interval - INTEGER | |
926 | Number of seconds to wait between Router Solicitations. | |
927 | Default: 4 | |
928 | ||
929 | router_solicitations - INTEGER | |
930 | Number of Router Solicitations to send until assuming no | |
931 | routers are present. | |
932 | Default: 3 | |
933 | ||
934 | use_tempaddr - INTEGER | |
935 | Preference for Privacy Extensions (RFC3041). | |
936 | <= 0 : disable Privacy Extensions | |
937 | == 1 : enable Privacy Extensions, but prefer public | |
938 | addresses over temporary addresses. | |
939 | > 1 : enable Privacy Extensions and prefer temporary | |
940 | addresses over public addresses. | |
941 | Default: 0 (for most devices) | |
942 | -1 (for point-to-point devices and loopback devices) | |
943 | ||
944 | temp_valid_lft - INTEGER | |
945 | valid lifetime (in seconds) for temporary addresses. | |
946 | Default: 604800 (7 days) | |
947 | ||
948 | temp_prefered_lft - INTEGER | |
949 | Preferred lifetime (in seconds) for temporary addresses. | |
950 | Default: 86400 (1 day) | |
951 | ||
952 | max_desync_factor - INTEGER | |
953 | Maximum value for DESYNC_FACTOR, which is a random value | |
954 | that ensures that clients don't synchronize with each | |
955 | other and generate new addresses at exactly the same time. | |
956 | value is in seconds. | |
957 | Default: 600 | |
958 | ||
959 | regen_max_retry - INTEGER | |
960 | Number of attempts before give up attempting to generate | |
961 | valid temporary addresses. | |
962 | Default: 5 | |
963 | ||
964 | max_addresses - INTEGER | |
965 | Number of maximum addresses per interface. 0 disables limitation. | |
966 | It is recommended not set too large value (or 0) because it would | |
967 | be too easy way to crash kernel to allow to create too much of | |
968 | autoconfigured addresses. | |
969 | Default: 16 | |
970 | ||
971 | icmp/*: | |
972 | ratelimit - INTEGER | |
973 | Limit the maximal rates for sending ICMPv6 packets. | |
974 | 0 to disable any limiting, otherwise the maximal rate in jiffies(1) | |
975 | Default: 100 | |
976 | ||
977 | ||
978 | IPv6 Update by: | |
979 | Pekka Savola <pekkas@netcore.fi> | |
980 | YOSHIFUJI Hideaki / USAGI Project <yoshfuji@linux-ipv6.org> | |
981 | ||
982 | ||
983 | /proc/sys/net/bridge/* Variables: | |
984 | ||
985 | bridge-nf-call-arptables - BOOLEAN | |
986 | 1 : pass bridged ARP traffic to arptables' FORWARD chain. | |
987 | 0 : disable this. | |
988 | Default: 1 | |
989 | ||
990 | bridge-nf-call-iptables - BOOLEAN | |
991 | 1 : pass bridged IPv4 traffic to iptables' chains. | |
992 | 0 : disable this. | |
993 | Default: 1 | |
994 | ||
995 | bridge-nf-call-ip6tables - BOOLEAN | |
996 | 1 : pass bridged IPv6 traffic to ip6tables' chains. | |
997 | 0 : disable this. | |
998 | Default: 1 | |
999 | ||
1000 | bridge-nf-filter-vlan-tagged - BOOLEAN | |
1001 | 1 : pass bridged vlan-tagged ARP/IP traffic to arptables/iptables. | |
1002 | 0 : disable this. | |
1003 | Default: 1 | |
1004 | ||
1005 | ||
1006 | UNDOCUMENTED: | |
1007 | ||
1008 | dev_weight FIXME | |
1009 | discovery_slots FIXME | |
1010 | discovery_timeout FIXME | |
1011 | fast_poll_increase FIXME | |
1012 | ip6_queue_maxlen FIXME | |
1013 | lap_keepalive_time FIXME | |
1014 | lo_cong FIXME | |
1015 | max_baud_rate FIXME | |
1016 | max_dgram_qlen FIXME | |
1017 | max_noreply_time FIXME | |
1018 | max_tx_data_size FIXME | |
1019 | max_tx_window FIXME | |
1020 | min_tx_turn_time FIXME | |
1021 | mod_cong FIXME | |
1022 | no_cong FIXME | |
1023 | no_cong_thresh FIXME | |
1024 | slot_timeout FIXME | |
1025 | warn_noreply_time FIXME | |
1026 |