Commit | Line | Data |
---|---|---|
1da177e4 LT |
1 | /proc/sys/net/ipv4/* Variables: |
2 | ||
3 | ip_forward - BOOLEAN | |
4 | 0 - disabled (default) | |
e18f5feb | 5 | not 0 - enabled |
1da177e4 LT |
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 | |
cc6f02dd ED |
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) | |
1da177e4 LT |
17 | |
18 | ip_no_pmtu_disc - BOOLEAN | |
19 | Disable Path MTU Discovery. | |
20 | default FALSE | |
21 | ||
22 | min_pmtu - INTEGER | |
20db93c3 | 23 | default 552 - minimum discovered Path MTU |
1da177e4 | 24 | |
6fa3eb70 S |
25 | fwmark_reflect - BOOLEAN |
26 | Controls the fwmark of kernel-generated IPv4 reply packets that are not | |
27 | associated with a socket for example, TCP RSTs or ICMP echo replies). | |
28 | If unset, these packets have a fwmark of zero. If set, they have the | |
29 | fwmark of the packet they are replying to. Similarly affects the fwmark | |
30 | used by internal routing lookups triggered by incoming packets, such as | |
31 | the ones used for Path MTU Discovery. | |
32 | Default: 0 | |
33 | ||
cbaf087a BG |
34 | route/max_size - INTEGER |
35 | Maximum number of routes allowed in the kernel. Increase | |
36 | this when using large numbers of interfaces and/or routes. | |
37 | ||
2724680b YH |
38 | neigh/default/gc_thresh1 - INTEGER |
39 | Minimum number of entries to keep. Garbage collector will not | |
40 | purge entries if there are fewer than this number. | |
b66c66dc | 41 | Default: 128 |
2724680b | 42 | |
cbaf087a BG |
43 | neigh/default/gc_thresh3 - INTEGER |
44 | Maximum number of neighbor entries allowed. Increase this | |
45 | when using large numbers of interfaces and when communicating | |
46 | with large numbers of directly-connected peers. | |
cc868028 | 47 | Default: 1024 |
cbaf087a | 48 | |
8b5c171b ED |
49 | neigh/default/unres_qlen_bytes - INTEGER |
50 | The maximum number of bytes which may be used by packets | |
51 | queued for each unresolved address by other network layers. | |
52 | (added in linux 3.3) | |
3b09adcb | 53 | Setting negative value is meaningless and will return error. |
cc868028 | 54 | Default: 65536 Bytes(64KB) |
8b5c171b ED |
55 | |
56 | neigh/default/unres_qlen - INTEGER | |
57 | The maximum number of packets which may be queued for each | |
58 | unresolved address by other network layers. | |
59 | (deprecated in linux 3.3) : use unres_qlen_bytes instead. | |
cc868028 | 60 | Prior to linux 3.3, the default value is 3 which may cause |
5d248c49 | 61 | unexpected packet loss. The current default value is calculated |
cc868028 SW |
62 | according to default value of unres_qlen_bytes and true size of |
63 | packet. | |
64 | Default: 31 | |
8b5c171b | 65 | |
1da177e4 LT |
66 | mtu_expires - INTEGER |
67 | Time, in seconds, that cached PMTU information is kept. | |
68 | ||
69 | min_adv_mss - INTEGER | |
70 | The advertised MSS depends on the first hop route MTU, but will | |
71 | never be lower than this setting. | |
72 | ||
73 | IP Fragmentation: | |
74 | ||
75 | ipfrag_high_thresh - INTEGER | |
e18f5feb | 76 | Maximum memory used to reassemble IP fragments. When |
1da177e4 LT |
77 | ipfrag_high_thresh bytes of memory is allocated for this purpose, |
78 | the fragment handler will toss packets until ipfrag_low_thresh | |
79 | is reached. | |
e18f5feb | 80 | |
1da177e4 | 81 | ipfrag_low_thresh - INTEGER |
e18f5feb | 82 | See ipfrag_high_thresh |
1da177e4 LT |
83 | |
84 | ipfrag_time - INTEGER | |
e18f5feb | 85 | Time in seconds to keep an IP fragment in memory. |
1da177e4 LT |
86 | |
87 | ipfrag_secret_interval - INTEGER | |
e18f5feb | 88 | Regeneration interval (in seconds) of the hash secret (or lifetime |
1da177e4 LT |
89 | for the hash secret) for IP fragments. |
90 | Default: 600 | |
91 | ||
89cee8b1 | 92 | ipfrag_max_dist - INTEGER |
e18f5feb JDB |
93 | ipfrag_max_dist is a non-negative integer value which defines the |
94 | maximum "disorder" which is allowed among fragments which share a | |
95 | common IP source address. Note that reordering of packets is | |
96 | not unusual, but if a large number of fragments arrive from a source | |
97 | IP address while a particular fragment queue remains incomplete, it | |
98 | probably indicates that one or more fragments belonging to that queue | |
99 | have been lost. When ipfrag_max_dist is positive, an additional check | |
100 | is done on fragments before they are added to a reassembly queue - if | |
101 | ipfrag_max_dist (or more) fragments have arrived from a particular IP | |
102 | address between additions to any IP fragment queue using that source | |
103 | address, it's presumed that one or more fragments in the queue are | |
104 | lost. The existing fragment queue will be dropped, and a new one | |
89cee8b1 HX |
105 | started. An ipfrag_max_dist value of zero disables this check. |
106 | ||
107 | Using a very small value, e.g. 1 or 2, for ipfrag_max_dist can | |
108 | result in unnecessarily dropping fragment queues when normal | |
e18f5feb JDB |
109 | reordering of packets occurs, which could lead to poor application |
110 | performance. Using a very large value, e.g. 50000, increases the | |
111 | likelihood of incorrectly reassembling IP fragments that originate | |
89cee8b1 HX |
112 | from different IP datagrams, which could result in data corruption. |
113 | Default: 64 | |
114 | ||
1da177e4 LT |
115 | INET peer storage: |
116 | ||
117 | inet_peer_threshold - INTEGER | |
e18f5feb | 118 | The approximate size of the storage. Starting from this threshold |
1da177e4 LT |
119 | entries will be thrown aggressively. This threshold also determines |
120 | entries' time-to-live and time intervals between garbage collection | |
121 | passes. More entries, less time-to-live, less GC interval. | |
122 | ||
123 | inet_peer_minttl - INTEGER | |
124 | Minimum time-to-live of entries. Should be enough to cover fragment | |
125 | time-to-live on the reassembling side. This minimum time-to-live is | |
126 | guaranteed if the pool size is less than inet_peer_threshold. | |
77a538d5 | 127 | Measured in seconds. |
1da177e4 LT |
128 | |
129 | inet_peer_maxttl - INTEGER | |
130 | Maximum time-to-live of entries. Unused entries will expire after | |
131 | this period of time if there is no memory pressure on the pool (i.e. | |
132 | when the number of entries in the pool is very small). | |
77a538d5 | 133 | Measured in seconds. |
1da177e4 | 134 | |
e18f5feb | 135 | TCP variables: |
1da177e4 | 136 | |
ef56e622 SH |
137 | somaxconn - INTEGER |
138 | Limit of socket listen() backlog, known in userspace as SOMAXCONN. | |
139 | Defaults to 128. See also tcp_max_syn_backlog for additional tuning | |
140 | for TCP sockets. | |
141 | ||
ef56e622 SH |
142 | tcp_abort_on_overflow - BOOLEAN |
143 | If listening service is too slow to accept new connections, | |
144 | reset them. Default state is FALSE. It means that if overflow | |
145 | occurred due to a burst, connection will recover. Enable this | |
146 | option _only_ if you are really sure that listening daemon | |
147 | cannot be tuned to accept connections faster. Enabling this | |
148 | option can harm clients of your server. | |
1da177e4 | 149 | |
ef56e622 SH |
150 | tcp_adv_win_scale - INTEGER |
151 | Count buffering overhead as bytes/2^tcp_adv_win_scale | |
152 | (if tcp_adv_win_scale > 0) or bytes-bytes/2^(-tcp_adv_win_scale), | |
153 | if it is <= 0. | |
0147fc05 | 154 | Possible values are [-31, 31], inclusive. |
b49960a0 | 155 | Default: 1 |
1da177e4 | 156 | |
ef56e622 SH |
157 | tcp_allowed_congestion_control - STRING |
158 | Show/set the congestion control choices available to non-privileged | |
159 | processes. The list is a subset of those listed in | |
160 | tcp_available_congestion_control. | |
161 | Default is "reno" and the default setting (tcp_congestion_control). | |
1da177e4 | 162 | |
ef56e622 SH |
163 | tcp_app_win - INTEGER |
164 | Reserve max(window/2^tcp_app_win, mss) of window for application | |
165 | buffer. Value 0 is special, it means that nothing is reserved. | |
166 | Default: 31 | |
1da177e4 | 167 | |
ef56e622 SH |
168 | tcp_available_congestion_control - STRING |
169 | Shows the available congestion control choices that are registered. | |
170 | More congestion control algorithms may be available as modules, | |
171 | but not loaded. | |
1da177e4 | 172 | |
71599cd1 | 173 | tcp_base_mss - INTEGER |
4edc2f34 SH |
174 | The initial value of search_low to be used by the packetization layer |
175 | Path MTU discovery (MTU probing). If MTU probing is enabled, | |
176 | this is the initial MSS used by the connection. | |
71599cd1 | 177 | |
ef56e622 SH |
178 | tcp_congestion_control - STRING |
179 | Set the congestion control algorithm to be used for new | |
180 | connections. The algorithm "reno" is always available, but | |
181 | additional choices may be available based on kernel configuration. | |
182 | Default is set as part of kernel configuration. | |
d8a6e65f ED |
183 | For passive connections, the listener congestion control choice |
184 | is inherited. | |
185 | [see setsockopt(listenfd, SOL_TCP, TCP_CONGESTION, "name" ...) ] | |
1da177e4 | 186 | |
ef56e622 SH |
187 | tcp_dsack - BOOLEAN |
188 | Allows TCP to send "duplicate" SACKs. | |
1da177e4 | 189 | |
eed530b6 YC |
190 | tcp_early_retrans - INTEGER |
191 | Enable Early Retransmit (ER), per RFC 5827. ER lowers the threshold | |
192 | for triggering fast retransmit when the amount of outstanding data is | |
193 | small and when no previously unsent data can be transmitted (such | |
6ba8a3b1 ND |
194 | that limited transmit could be used). Also controls the use of |
195 | Tail loss probe (TLP) that converts RTOs occuring due to tail | |
196 | losses into fast recovery (draft-dukkipati-tcpm-tcp-loss-probe-01). | |
eed530b6 YC |
197 | Possible values: |
198 | 0 disables ER | |
199 | 1 enables ER | |
200 | 2 enables ER but delays fast recovery and fast retransmit | |
201 | by a fourth of RTT. This mitigates connection falsely | |
202 | recovers when network has a small degree of reordering | |
203 | (less than 3 packets). | |
6ba8a3b1 ND |
204 | 3 enables delayed ER and TLP. |
205 | 4 enables TLP only. | |
206 | Default: 3 | |
eed530b6 | 207 | |
34a6ef38 | 208 | tcp_ecn - INTEGER |
7e3a2dc5 RJ |
209 | Control use of Explicit Congestion Notification (ECN) by TCP. |
210 | ECN is used only when both ends of the TCP connection indicate | |
211 | support for it. This feature is useful in avoiding losses due | |
212 | to congestion by allowing supporting routers to signal | |
213 | congestion before having to drop packets. | |
255cac91 | 214 | Possible values are: |
7e3a2dc5 | 215 | 0 Disable ECN. Neither initiate nor accept ECN. |
3d55b323 VS |
216 | 1 Enable ECN when requested by incoming connections and |
217 | also request ECN on outgoing connection attempts. | |
218 | 2 Enable ECN when requested by incoming connections | |
7e3a2dc5 | 219 | but do not request ECN on outgoing connections. |
255cac91 | 220 | Default: 2 |
ef56e622 SH |
221 | |
222 | tcp_fack - BOOLEAN | |
223 | Enable FACK congestion avoidance and fast retransmission. | |
224 | The value is not used, if tcp_sack is not enabled. | |
1da177e4 LT |
225 | |
226 | tcp_fin_timeout - INTEGER | |
d825da2e RJ |
227 | The length of time an orphaned (no longer referenced by any |
228 | application) connection will remain in the FIN_WAIT_2 state | |
229 | before it is aborted at the local end. While a perfectly | |
230 | valid "receive only" state for an un-orphaned connection, an | |
231 | orphaned connection in FIN_WAIT_2 state could otherwise wait | |
232 | forever for the remote to close its end of the connection. | |
233 | Cf. tcp_max_orphans | |
234 | Default: 60 seconds | |
1da177e4 | 235 | |
89808060 | 236 | tcp_frto - INTEGER |
e33099f9 | 237 | Enables Forward RTO-Recovery (F-RTO) defined in RFC5682. |
cd99889c | 238 | F-RTO is an enhanced recovery algorithm for TCP retransmission |
e33099f9 YC |
239 | timeouts. It is particularly beneficial in networks where the |
240 | RTT fluctuates (e.g., wireless). F-RTO is sender-side only | |
241 | modification. It does not require any support from the peer. | |
242 | ||
243 | By default it's enabled with a non-zero value. 0 disables F-RTO. | |
1da177e4 | 244 | |
ef56e622 SH |
245 | tcp_keepalive_time - INTEGER |
246 | How often TCP sends out keepalive messages when keepalive is enabled. | |
247 | Default: 2hours. | |
1da177e4 | 248 | |
ef56e622 SH |
249 | tcp_keepalive_probes - INTEGER |
250 | How many keepalive probes TCP sends out, until it decides that the | |
251 | connection is broken. Default value: 9. | |
252 | ||
253 | tcp_keepalive_intvl - INTEGER | |
254 | How frequently the probes are send out. Multiplied by | |
255 | tcp_keepalive_probes it is time to kill not responding connection, | |
256 | after probes started. Default value: 75sec i.e. connection | |
257 | will be aborted after ~11 minutes of retries. | |
258 | ||
259 | tcp_low_latency - BOOLEAN | |
260 | If set, the TCP stack makes decisions that prefer lower | |
261 | latency as opposed to higher throughput. By default, this | |
262 | option is not set meaning that higher throughput is preferred. | |
263 | An example of an application where this default should be | |
264 | changed would be a Beowulf compute cluster. | |
265 | Default: 0 | |
1da177e4 LT |
266 | |
267 | tcp_max_orphans - INTEGER | |
268 | Maximal number of TCP sockets not attached to any user file handle, | |
269 | held by system. If this number is exceeded orphaned connections are | |
270 | reset immediately and warning is printed. This limit exists | |
271 | only to prevent simple DoS attacks, you _must_ not rely on this | |
272 | or lower the limit artificially, but rather increase it | |
273 | (probably, after increasing installed memory), | |
274 | if network conditions require more than default value, | |
275 | and tune network services to linger and kill such states | |
276 | more aggressively. Let me to remind again: each orphan eats | |
277 | up to ~64K of unswappable memory. | |
278 | ||
81146ec1 IJ |
279 | tcp_max_ssthresh - INTEGER |
280 | Limited Slow-Start for TCP with large congestion windows (cwnd) defined in | |
281 | RFC3742. Limited slow-start is a mechanism to limit growth of the cwnd | |
282 | on the region where cwnd is larger than tcp_max_ssthresh. TCP increases cwnd | |
283 | by at most tcp_max_ssthresh segments, and by at least tcp_max_ssthresh/2 | |
284 | segments per RTT when the cwnd is above tcp_max_ssthresh. | |
285 | If TCP connection increased cwnd to thousands (or tens of thousands) segments, | |
286 | and thousands of packets were being dropped during slow-start, you can set | |
287 | tcp_max_ssthresh to improve performance for new TCP connection. | |
288 | Default: 0 (off) | |
289 | ||
1da177e4 | 290 | tcp_max_syn_backlog - INTEGER |
99b53bdd PP |
291 | Maximal number of remembered connection requests, which have not |
292 | received an acknowledgment from connecting client. | |
293 | The minimal value is 128 for low memory machines, and it will | |
294 | increase in proportion to the memory of machine. | |
295 | If server suffers from overload, try increasing this number. | |
1da177e4 | 296 | |
ef56e622 SH |
297 | tcp_max_tw_buckets - INTEGER |
298 | Maximal number of timewait sockets held by system simultaneously. | |
299 | If this number is exceeded time-wait socket is immediately destroyed | |
300 | and warning is printed. This limit exists only to prevent | |
301 | simple DoS attacks, you _must_ not lower the limit artificially, | |
302 | but rather increase it (probably, after increasing installed memory), | |
303 | if network conditions require more than default value. | |
1da177e4 | 304 | |
ef56e622 SH |
305 | tcp_mem - vector of 3 INTEGERs: min, pressure, max |
306 | min: below this number of pages TCP is not bothered about its | |
307 | memory appetite. | |
1da177e4 | 308 | |
ef56e622 SH |
309 | pressure: when amount of memory allocated by TCP exceeds this number |
310 | of pages, TCP moderates its memory consumption and enters memory | |
311 | pressure mode, which is exited when memory consumption falls | |
312 | under "min". | |
1da177e4 | 313 | |
ef56e622 | 314 | max: number of pages allowed for queueing by all TCP sockets. |
1da177e4 | 315 | |
ef56e622 SH |
316 | Defaults are calculated at boot time from amount of available |
317 | memory. | |
1da177e4 | 318 | |
71599cd1 | 319 | tcp_moderate_rcvbuf - BOOLEAN |
4edc2f34 | 320 | If set, TCP performs receive buffer auto-tuning, attempting to |
71599cd1 JH |
321 | automatically size the buffer (no greater than tcp_rmem[2]) to |
322 | match the size required by the path for full throughput. Enabled by | |
323 | default. | |
324 | ||
325 | tcp_mtu_probing - INTEGER | |
326 | Controls TCP Packetization-Layer Path MTU Discovery. Takes three | |
327 | values: | |
328 | 0 - Disabled | |
329 | 1 - Disabled by default, enabled when an ICMP black hole detected | |
330 | 2 - Always enabled, use initial MSS of tcp_base_mss. | |
331 | ||
332 | tcp_no_metrics_save - BOOLEAN | |
333 | By default, TCP saves various connection metrics in the route cache | |
334 | when the connection closes, so that connections established in the | |
335 | near future can use these to set initial conditions. Usually, this | |
336 | increases overall performance, but may sometimes cause performance | |
0f035b8e | 337 | degradation. If set, TCP will not cache metrics on closing |
71599cd1 JH |
338 | connections. |
339 | ||
ef56e622 | 340 | tcp_orphan_retries - INTEGER |
5d789229 DL |
341 | This value influences the timeout of a locally closed TCP connection, |
342 | when RTO retransmissions remain unacknowledged. | |
343 | See tcp_retries2 for more details. | |
344 | ||
06b8fc5d | 345 | The default value is 8. |
5d789229 | 346 | If your machine is a loaded WEB server, |
ef56e622 SH |
347 | you should think about lowering this value, such sockets |
348 | may consume significant resources. Cf. tcp_max_orphans. | |
1da177e4 LT |
349 | |
350 | tcp_reordering - INTEGER | |
351 | Maximal reordering of packets in a TCP stream. | |
e18f5feb | 352 | Default: 3 |
1da177e4 LT |
353 | |
354 | tcp_retrans_collapse - BOOLEAN | |
355 | Bug-to-bug compatibility with some broken printers. | |
356 | On retransmit try to send bigger packets to work around bugs in | |
357 | certain TCP stacks. | |
358 | ||
ef56e622 | 359 | tcp_retries1 - INTEGER |
5d789229 DL |
360 | This value influences the time, after which TCP decides, that |
361 | something is wrong due to unacknowledged RTO retransmissions, | |
362 | and reports this suspicion to the network layer. | |
363 | See tcp_retries2 for more details. | |
364 | ||
365 | RFC 1122 recommends at least 3 retransmissions, which is the | |
366 | default. | |
1da177e4 | 367 | |
ef56e622 | 368 | tcp_retries2 - INTEGER |
5d789229 DL |
369 | This value influences the timeout of an alive TCP connection, |
370 | when RTO retransmissions remain unacknowledged. | |
371 | Given a value of N, a hypothetical TCP connection following | |
372 | exponential backoff with an initial RTO of TCP_RTO_MIN would | |
373 | retransmit N times before killing the connection at the (N+1)th RTO. | |
374 | ||
375 | The default value of 15 yields a hypothetical timeout of 924.6 | |
376 | seconds and is a lower bound for the effective timeout. | |
377 | TCP will effectively time out at the first RTO which exceeds the | |
378 | hypothetical timeout. | |
379 | ||
380 | RFC 1122 recommends at least 100 seconds for the timeout, | |
381 | which corresponds to a value of at least 8. | |
1da177e4 | 382 | |
ef56e622 SH |
383 | tcp_rfc1337 - BOOLEAN |
384 | If set, the TCP stack behaves conforming to RFC1337. If unset, | |
385 | we are not conforming to RFC, but prevent TCP TIME_WAIT | |
386 | assassination. | |
387 | Default: 0 | |
1da177e4 LT |
388 | |
389 | tcp_rmem - vector of 3 INTEGERs: min, default, max | |
390 | min: Minimal size of receive buffer used by TCP sockets. | |
391 | It is guaranteed to each TCP socket, even under moderate memory | |
392 | pressure. | |
6539fefd | 393 | Default: 1 page |
1da177e4 | 394 | |
53025f5e | 395 | default: initial size of receive buffer used by TCP sockets. |
1da177e4 LT |
396 | This value overrides net.core.rmem_default used by other protocols. |
397 | Default: 87380 bytes. This value results in window of 65535 with | |
398 | default setting of tcp_adv_win_scale and tcp_app_win:0 and a bit | |
399 | less for default tcp_app_win. See below about these variables. | |
400 | ||
401 | max: maximal size of receive buffer allowed for automatically | |
402 | selected receiver buffers for TCP socket. This value does not override | |
53025f5e BF |
403 | net.core.rmem_max. Calling setsockopt() with SO_RCVBUF disables |
404 | automatic tuning of that socket's receive buffer size, in which | |
405 | case this value is ignored. | |
b49960a0 | 406 | Default: between 87380B and 6MB, depending on RAM size. |
1da177e4 | 407 | |
ef56e622 SH |
408 | tcp_sack - BOOLEAN |
409 | Enable select acknowledgments (SACKS). | |
1da177e4 | 410 | |
ef56e622 SH |
411 | tcp_slow_start_after_idle - BOOLEAN |
412 | If set, provide RFC2861 behavior and time out the congestion | |
413 | window after an idle period. An idle period is defined at | |
414 | the current RTO. If unset, the congestion window will not | |
415 | be timed out after an idle period. | |
416 | Default: 1 | |
1da177e4 | 417 | |
ef56e622 | 418 | tcp_stdurg - BOOLEAN |
4edc2f34 | 419 | Use the Host requirements interpretation of the TCP urgent pointer field. |
ef56e622 SH |
420 | Most hosts use the older BSD interpretation, so if you turn this on |
421 | Linux might not communicate correctly with them. | |
422 | Default: FALSE | |
1da177e4 | 423 | |
ef56e622 SH |
424 | tcp_synack_retries - INTEGER |
425 | Number of times SYNACKs for a passive TCP connection attempt will | |
426 | be retransmitted. Should not be higher than 255. Default value | |
6c9ff979 AB |
427 | is 5, which corresponds to 31seconds till the last retransmission |
428 | with the current initial RTO of 1second. With this the final timeout | |
429 | for a passive TCP connection will happen after 63seconds. | |
1da177e4 | 430 | |
ef56e622 | 431 | tcp_syncookies - BOOLEAN |
a3c910d2 | 432 | Only valid when the kernel was compiled with CONFIG_SYN_COOKIES |
ef56e622 | 433 | Send out syncookies when the syn backlog queue of a socket |
4edc2f34 | 434 | overflows. This is to prevent against the common 'SYN flood attack' |
a3c910d2 | 435 | Default: 1 |
1da177e4 | 436 | |
ef56e622 SH |
437 | Note, that syncookies is fallback facility. |
438 | It MUST NOT be used to help highly loaded servers to stand | |
4edc2f34 | 439 | against legal connection rate. If you see SYN flood warnings |
ef56e622 SH |
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. | |
1da177e4 | 444 | |
ef56e622 SH |
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 | |
4edc2f34 | 449 | SYN flood warnings in logs not being really flooded, your server |
ef56e622 | 450 | is seriously misconfigured. |
1da177e4 | 451 | |
cf60af03 YC |
452 | tcp_fastopen - INTEGER |
453 | Enable TCP Fast Open feature (draft-ietf-tcpm-fastopen) to send data | |
454 | in the opening SYN packet. To use this feature, the client application | |
10467163 JC |
455 | must use sendmsg() or sendto() with MSG_FASTOPEN flag rather than |
456 | connect() to perform a TCP handshake automatically. | |
457 | ||
458 | The values (bitmap) are | |
459 | 1: Enables sending data in the opening SYN on the client. | |
460 | 2: Enables TCP Fast Open on the server side, i.e., allowing data in | |
461 | a SYN packet to be accepted and passed to the application before | |
462 | 3-way hand shake finishes. | |
463 | 4: Send data in the opening SYN regardless of cookie availability and | |
464 | without a cookie option. | |
465 | 0x100: Accept SYN data w/o validating the cookie. | |
466 | 0x200: Accept data-in-SYN w/o any cookie option present. | |
467 | 0x400/0x800: Enable Fast Open on all listeners regardless of the | |
468 | TCP_FASTOPEN socket option. The two different flags designate two | |
469 | different ways of setting max_qlen without the TCP_FASTOPEN socket | |
470 | option. | |
cf60af03 YC |
471 | |
472 | Default: 0 | |
473 | ||
10467163 JC |
474 | Note that the client & server side Fast Open flags (1 and 2 |
475 | respectively) must be also enabled before the rest of flags can take | |
476 | effect. | |
477 | ||
478 | See include/net/tcp.h and the code for more details. | |
479 | ||
6fa3eb70 S |
480 | tcp_fwmark_accept - BOOLEAN |
481 | If set, incoming connections to listening sockets that do not have a | |
482 | socket mark will set the mark of the accepting socket to the fwmark of | |
483 | the incoming SYN packet. This will cause all packets on that connection | |
484 | (starting from the first SYNACK) to be sent with that fwmark. The | |
485 | listening socket's mark is unchanged. Listening sockets that already | |
486 | have a fwmark set via setsockopt(SOL_SOCKET, SO_MARK, ...) are | |
487 | unaffected. | |
488 | Default: 0 | |
489 | ||
ef56e622 SH |
490 | tcp_syn_retries - INTEGER |
491 | Number of times initial SYNs for an active TCP connection attempt | |
492 | will be retransmitted. Should not be higher than 255. Default value | |
3b09adcb | 493 | is 6, which corresponds to 63seconds till the last retransmission |
6c9ff979 AB |
494 | with the current initial RTO of 1second. With this the final timeout |
495 | for an active TCP connection attempt will happen after 127seconds. | |
ef56e622 SH |
496 | |
497 | tcp_timestamps - BOOLEAN | |
498 | Enable timestamps as defined in RFC1323. | |
1da177e4 | 499 | |
5e25ba50 ED |
500 | tcp_min_tso_segs - INTEGER |
501 | Minimal number of segments per TSO frame. | |
502 | Since linux-3.12, TCP does an automatic sizing of TSO frames, | |
503 | depending on flow rate, instead of filling 64Kbytes packets. | |
504 | For specific usages, it's possible to force TCP to build big | |
505 | TSO frames. Note that TCP stack might split too big TSO packets | |
506 | if available window is too small. | |
507 | Default: 2 | |
508 | ||
1da177e4 | 509 | tcp_tso_win_divisor - INTEGER |
ef56e622 SH |
510 | This allows control over what percentage of the congestion window |
511 | can be consumed by a single TSO frame. | |
512 | The setting of this parameter is a choice between burstiness and | |
513 | building larger TSO frames. | |
514 | Default: 3 | |
1da177e4 | 515 | |
ef56e622 SH |
516 | tcp_tw_recycle - BOOLEAN |
517 | Enable fast recycling TIME-WAIT sockets. Default value is 0. | |
518 | It should not be changed without advice/request of technical | |
519 | experts. | |
1da177e4 | 520 | |
ef56e622 SH |
521 | tcp_tw_reuse - BOOLEAN |
522 | Allow to reuse TIME-WAIT sockets for new connections when it is | |
523 | safe from protocol viewpoint. Default value is 0. | |
524 | It should not be changed without advice/request of technical | |
525 | experts. | |
ce7bc3bf | 526 | |
ef56e622 SH |
527 | tcp_window_scaling - BOOLEAN |
528 | Enable window scaling as defined in RFC1323. | |
3ff825b2 | 529 | |
ef56e622 | 530 | tcp_wmem - vector of 3 INTEGERs: min, default, max |
53025f5e | 531 | min: Amount of memory reserved for send buffers for TCP sockets. |
ef56e622 | 532 | Each TCP socket has rights to use it due to fact of its birth. |
6539fefd | 533 | Default: 1 page |
9d7bcfc6 | 534 | |
53025f5e BF |
535 | default: initial size of send buffer used by TCP sockets. This |
536 | value overrides net.core.wmem_default used by other protocols. | |
537 | It is usually lower than net.core.wmem_default. | |
ef56e622 SH |
538 | Default: 16K |
539 | ||
53025f5e BF |
540 | max: Maximal amount of memory allowed for automatically tuned |
541 | send buffers for TCP sockets. This value does not override | |
542 | net.core.wmem_max. Calling setsockopt() with SO_SNDBUF disables | |
543 | automatic tuning of that socket's send buffer size, in which case | |
544 | this value is ignored. | |
545 | Default: between 64K and 4MB, depending on RAM size. | |
1da177e4 | 546 | |
15d99e02 RJ |
547 | tcp_workaround_signed_windows - BOOLEAN |
548 | If set, assume no receipt of a window scaling option means the | |
549 | remote TCP is broken and treats the window as a signed quantity. | |
550 | If unset, assume the remote TCP is not broken even if we do | |
551 | not receive a window scaling option from them. | |
552 | Default: 0 | |
553 | ||
72d0b7a8 CL |
554 | tcp_dma_copybreak - INTEGER |
555 | Lower limit, in bytes, of the size of socket reads that will be | |
556 | offloaded to a DMA copy engine, if one is present in the system | |
557 | and CONFIG_NET_DMA is enabled. | |
558 | Default: 4096 | |
559 | ||
36e31b0a AP |
560 | tcp_thin_linear_timeouts - BOOLEAN |
561 | Enable dynamic triggering of linear timeouts for thin streams. | |
562 | If set, a check is performed upon retransmission by timeout to | |
563 | determine if the stream is thin (less than 4 packets in flight). | |
564 | As long as the stream is found to be thin, up to 6 linear | |
565 | timeouts may be performed before exponential backoff mode is | |
566 | initiated. This improves retransmission latency for | |
567 | non-aggressive thin streams, often found to be time-dependent. | |
568 | For more information on thin streams, see | |
569 | Documentation/networking/tcp-thin.txt | |
570 | Default: 0 | |
571 | ||
7e380175 AP |
572 | tcp_thin_dupack - BOOLEAN |
573 | Enable dynamic triggering of retransmissions after one dupACK | |
574 | for thin streams. If set, a check is performed upon reception | |
575 | of a dupACK to determine if the stream is thin (less than 4 | |
576 | packets in flight). As long as the stream is found to be thin, | |
577 | data is retransmitted on the first received dupACK. This | |
578 | improves retransmission latency for non-aggressive thin | |
579 | streams, often found to be time-dependent. | |
580 | For more information on thin streams, see | |
581 | Documentation/networking/tcp-thin.txt | |
582 | Default: 0 | |
583 | ||
46d3ceab ED |
584 | tcp_limit_output_bytes - INTEGER |
585 | Controls TCP Small Queue limit per tcp socket. | |
586 | TCP bulk sender tends to increase packets in flight until it | |
587 | gets losses notifications. With SNDBUF autotuning, this can | |
588 | result in a large amount of packets queued in qdisc/device | |
589 | on the local machine, hurting latency of other flows, for | |
590 | typical pfifo_fast qdiscs. | |
591 | tcp_limit_output_bytes limits the number of bytes on qdisc | |
592 | or device to reduce artificial RTT/cwnd and reduce bufferbloat. | |
46d3ceab ED |
593 | Default: 131072 |
594 | ||
282f23c6 ED |
595 | tcp_challenge_ack_limit - INTEGER |
596 | Limits number of Challenge ACK sent per second, as recommended | |
597 | in RFC 5961 (Improving TCP's Robustness to Blind In-Window Attacks) | |
598 | Default: 100 | |
599 | ||
95766fff HA |
600 | UDP variables: |
601 | ||
602 | udp_mem - vector of 3 INTEGERs: min, pressure, max | |
603 | Number of pages allowed for queueing by all UDP sockets. | |
604 | ||
605 | min: Below this number of pages UDP is not bothered about its | |
606 | memory appetite. When amount of memory allocated by UDP exceeds | |
607 | this number, UDP starts to moderate memory usage. | |
608 | ||
609 | pressure: This value was introduced to follow format of tcp_mem. | |
610 | ||
611 | max: Number of pages allowed for queueing by all UDP sockets. | |
612 | ||
613 | Default is calculated at boot time from amount of available memory. | |
614 | ||
615 | udp_rmem_min - INTEGER | |
616 | Minimal size of receive buffer used by UDP sockets in moderation. | |
617 | Each UDP socket is able to use the size for receiving data, even if | |
618 | total pages of UDP sockets exceed udp_mem pressure. The unit is byte. | |
6539fefd | 619 | Default: 1 page |
95766fff HA |
620 | |
621 | udp_wmem_min - INTEGER | |
622 | Minimal size of send buffer used by UDP sockets in moderation. | |
623 | Each UDP socket is able to use the size for sending data, even if | |
624 | total pages of UDP sockets exceed udp_mem pressure. The unit is byte. | |
6539fefd | 625 | Default: 1 page |
95766fff | 626 | |
8802f616 PM |
627 | CIPSOv4 Variables: |
628 | ||
629 | cipso_cache_enable - BOOLEAN | |
630 | If set, enable additions to and lookups from the CIPSO label mapping | |
631 | cache. If unset, additions are ignored and lookups always result in a | |
632 | miss. However, regardless of the setting the cache is still | |
633 | invalidated when required when means you can safely toggle this on and | |
634 | off and the cache will always be "safe". | |
635 | Default: 1 | |
636 | ||
637 | cipso_cache_bucket_size - INTEGER | |
638 | The CIPSO label cache consists of a fixed size hash table with each | |
639 | hash bucket containing a number of cache entries. This variable limits | |
640 | the number of entries in each hash bucket; the larger the value the | |
641 | more CIPSO label mappings that can be cached. When the number of | |
642 | entries in a given hash bucket reaches this limit adding new entries | |
643 | causes the oldest entry in the bucket to be removed to make room. | |
644 | Default: 10 | |
645 | ||
646 | cipso_rbm_optfmt - BOOLEAN | |
647 | Enable the "Optimized Tag 1 Format" as defined in section 3.4.2.6 of | |
648 | the CIPSO draft specification (see Documentation/netlabel for details). | |
649 | This means that when set the CIPSO tag will be padded with empty | |
650 | categories in order to make the packet data 32-bit aligned. | |
651 | Default: 0 | |
652 | ||
653 | cipso_rbm_structvalid - BOOLEAN | |
654 | If set, do a very strict check of the CIPSO option when | |
655 | ip_options_compile() is called. If unset, relax the checks done during | |
656 | ip_options_compile(). Either way is "safe" as errors are caught else | |
657 | where in the CIPSO processing code but setting this to 0 (False) should | |
658 | result in less work (i.e. it should be faster) but could cause problems | |
659 | with other implementations that require strict checking. | |
660 | Default: 0 | |
661 | ||
1da177e4 LT |
662 | IP Variables: |
663 | ||
664 | ip_local_port_range - 2 INTEGERS | |
665 | Defines the local port range that is used by TCP and UDP to | |
e18f5feb | 666 | choose the local port. The first number is the first, the |
5d6bd861 FLVC |
667 | second the last local port number. The default values are |
668 | 32768 and 61000 respectively. | |
1da177e4 | 669 | |
e3826f1e AW |
670 | ip_local_reserved_ports - list of comma separated ranges |
671 | Specify the ports which are reserved for known third-party | |
672 | applications. These ports will not be used by automatic port | |
673 | assignments (e.g. when calling connect() or bind() with port | |
674 | number 0). Explicit port allocation behavior is unchanged. | |
675 | ||
676 | The format used for both input and output is a comma separated | |
677 | list of ranges (e.g. "1,2-4,10-10" for ports 1, 2, 3, 4 and | |
678 | 10). Writing to the file will clear all previously reserved | |
679 | ports and update the current list with the one given in the | |
680 | input. | |
681 | ||
682 | Note that ip_local_port_range and ip_local_reserved_ports | |
683 | settings are independent and both are considered by the kernel | |
684 | when determining which ports are available for automatic port | |
685 | assignments. | |
686 | ||
687 | You can reserve ports which are not in the current | |
688 | ip_local_port_range, e.g.: | |
689 | ||
690 | $ cat /proc/sys/net/ipv4/ip_local_port_range | |
691 | 32000 61000 | |
692 | $ cat /proc/sys/net/ipv4/ip_local_reserved_ports | |
693 | 8080,9148 | |
694 | ||
695 | although this is redundant. However such a setting is useful | |
696 | if later the port range is changed to a value that will | |
697 | include the reserved ports. | |
698 | ||
699 | Default: Empty | |
700 | ||
1da177e4 LT |
701 | ip_nonlocal_bind - BOOLEAN |
702 | If set, allows processes to bind() to non-local IP addresses, | |
703 | which can be quite useful - but may break some applications. | |
704 | Default: 0 | |
705 | ||
706 | ip_dynaddr - BOOLEAN | |
707 | If set non-zero, enables support for dynamic addresses. | |
708 | If set to a non-zero value larger than 1, a kernel log | |
709 | message will be printed when dynamic address rewriting | |
710 | occurs. | |
711 | Default: 0 | |
712 | ||
713 | icmp_echo_ignore_all - BOOLEAN | |
7ce31246 DM |
714 | If set non-zero, then the kernel will ignore all ICMP ECHO |
715 | requests sent to it. | |
716 | Default: 0 | |
717 | ||
1da177e4 | 718 | icmp_echo_ignore_broadcasts - BOOLEAN |
7ce31246 DM |
719 | If set non-zero, then the kernel will ignore all ICMP ECHO and |
720 | TIMESTAMP requests sent to it via broadcast/multicast. | |
721 | Default: 1 | |
1da177e4 LT |
722 | |
723 | icmp_ratelimit - INTEGER | |
724 | Limit the maximal rates for sending ICMP packets whose type matches | |
725 | icmp_ratemask (see below) to specific targets. | |
6dbf4bca SH |
726 | 0 to disable any limiting, |
727 | otherwise the minimal space between responses in milliseconds. | |
728 | Default: 1000 | |
1da177e4 LT |
729 | |
730 | icmp_ratemask - INTEGER | |
731 | Mask made of ICMP types for which rates are being limited. | |
732 | Significant bits: IHGFEDCBA9876543210 | |
733 | Default mask: 0000001100000011000 (6168) | |
734 | ||
735 | Bit definitions (see include/linux/icmp.h): | |
736 | 0 Echo Reply | |
737 | 3 Destination Unreachable * | |
738 | 4 Source Quench * | |
739 | 5 Redirect | |
740 | 8 Echo Request | |
741 | B Time Exceeded * | |
742 | C Parameter Problem * | |
743 | D Timestamp Request | |
744 | E Timestamp Reply | |
745 | F Info Request | |
746 | G Info Reply | |
747 | H Address Mask Request | |
748 | I Address Mask Reply | |
749 | ||
750 | * These are rate limited by default (see default mask above) | |
751 | ||
752 | icmp_ignore_bogus_error_responses - BOOLEAN | |
753 | Some routers violate RFC1122 by sending bogus responses to broadcast | |
754 | frames. Such violations are normally logged via a kernel warning. | |
755 | If this is set to TRUE, the kernel will not give such warnings, which | |
756 | will avoid log file clutter. | |
757 | Default: FALSE | |
758 | ||
95f7daf1 H |
759 | icmp_errors_use_inbound_ifaddr - BOOLEAN |
760 | ||
761 | If zero, icmp error messages are sent with the primary address of | |
762 | the exiting interface. | |
e18f5feb | 763 | |
95f7daf1 H |
764 | If non-zero, the message will be sent with the primary address of |
765 | the interface that received the packet that caused the icmp error. | |
766 | This is the behaviour network many administrators will expect from | |
767 | a router. And it can make debugging complicated network layouts | |
e18f5feb | 768 | much easier. |
95f7daf1 H |
769 | |
770 | Note that if no primary address exists for the interface selected, | |
771 | then the primary address of the first non-loopback interface that | |
d6bc8ac9 | 772 | has one will be used regardless of this setting. |
95f7daf1 H |
773 | |
774 | Default: 0 | |
775 | ||
1da177e4 LT |
776 | igmp_max_memberships - INTEGER |
777 | Change the maximum number of multicast groups we can subscribe to. | |
778 | Default: 20 | |
779 | ||
d67ef35f JE |
780 | Theoretical maximum value is bounded by having to send a membership |
781 | report in a single datagram (i.e. the report can't span multiple | |
782 | datagrams, or risk confusing the switch and leaving groups you don't | |
783 | intend to). | |
1da177e4 | 784 | |
d67ef35f JE |
785 | The number of supported groups 'M' is bounded by the number of group |
786 | report entries you can fit into a single datagram of 65535 bytes. | |
787 | ||
788 | M = 65536-sizeof (ip header)/(sizeof(Group record)) | |
789 | ||
790 | Group records are variable length, with a minimum of 12 bytes. | |
791 | So net.ipv4.igmp_max_memberships should not be set higher than: | |
792 | ||
793 | (65536-24) / 12 = 5459 | |
794 | ||
795 | The value 5459 assumes no IP header options, so in practice | |
796 | this number may be lower. | |
797 | ||
798 | conf/interface/* changes special settings per interface (where | |
799 | "interface" is the name of your network interface) | |
800 | ||
801 | conf/all/* is special, changes the settings for all interfaces | |
1da177e4 LT |
802 | |
803 | log_martians - BOOLEAN | |
804 | Log packets with impossible addresses to kernel log. | |
805 | log_martians for the interface will be enabled if at least one of | |
806 | conf/{all,interface}/log_martians is set to TRUE, | |
807 | it will be disabled otherwise | |
808 | ||
809 | accept_redirects - BOOLEAN | |
810 | Accept ICMP redirect messages. | |
811 | accept_redirects for the interface will be enabled if: | |
e18f5feb JDB |
812 | - both conf/{all,interface}/accept_redirects are TRUE in the case |
813 | forwarding for the interface is enabled | |
1da177e4 | 814 | or |
e18f5feb JDB |
815 | - at least one of conf/{all,interface}/accept_redirects is TRUE in the |
816 | case forwarding for the interface is disabled | |
1da177e4 LT |
817 | accept_redirects for the interface will be disabled otherwise |
818 | default TRUE (host) | |
819 | FALSE (router) | |
820 | ||
821 | forwarding - BOOLEAN | |
822 | Enable IP forwarding on this interface. | |
823 | ||
824 | mc_forwarding - BOOLEAN | |
825 | Do multicast routing. The kernel needs to be compiled with CONFIG_MROUTE | |
826 | and a multicast routing daemon is required. | |
e18f5feb JDB |
827 | conf/all/mc_forwarding must also be set to TRUE to enable multicast |
828 | routing for the interface | |
1da177e4 LT |
829 | |
830 | medium_id - INTEGER | |
831 | Integer value used to differentiate the devices by the medium they | |
832 | are attached to. Two devices can have different id values when | |
833 | the broadcast packets are received only on one of them. | |
834 | The default value 0 means that the device is the only interface | |
835 | to its medium, value of -1 means that medium is not known. | |
e18f5feb | 836 | |
1da177e4 LT |
837 | Currently, it is used to change the proxy_arp behavior: |
838 | the proxy_arp feature is enabled for packets forwarded between | |
839 | two devices attached to different media. | |
840 | ||
841 | proxy_arp - BOOLEAN | |
842 | Do proxy arp. | |
843 | proxy_arp for the interface will be enabled if at least one of | |
844 | conf/{all,interface}/proxy_arp is set to TRUE, | |
845 | it will be disabled otherwise | |
846 | ||
65324144 JDB |
847 | proxy_arp_pvlan - BOOLEAN |
848 | Private VLAN proxy arp. | |
849 | Basically allow proxy arp replies back to the same interface | |
850 | (from which the ARP request/solicitation was received). | |
851 | ||
852 | This is done to support (ethernet) switch features, like RFC | |
853 | 3069, where the individual ports are NOT allowed to | |
854 | communicate with each other, but they are allowed to talk to | |
855 | the upstream router. As described in RFC 3069, it is possible | |
856 | to allow these hosts to communicate through the upstream | |
857 | router by proxy_arp'ing. Don't need to be used together with | |
858 | proxy_arp. | |
859 | ||
860 | This technology is known by different names: | |
861 | In RFC 3069 it is called VLAN Aggregation. | |
862 | Cisco and Allied Telesyn call it Private VLAN. | |
863 | Hewlett-Packard call it Source-Port filtering or port-isolation. | |
864 | Ericsson call it MAC-Forced Forwarding (RFC Draft). | |
865 | ||
1da177e4 LT |
866 | shared_media - BOOLEAN |
867 | Send(router) or accept(host) RFC1620 shared media redirects. | |
868 | Overrides ip_secure_redirects. | |
869 | shared_media for the interface will be enabled if at least one of | |
870 | conf/{all,interface}/shared_media is set to TRUE, | |
871 | it will be disabled otherwise | |
872 | default TRUE | |
873 | ||
874 | secure_redirects - BOOLEAN | |
875 | Accept ICMP redirect messages only for gateways, | |
876 | listed in default gateway list. | |
877 | secure_redirects for the interface will be enabled if at least one of | |
878 | conf/{all,interface}/secure_redirects is set to TRUE, | |
879 | it will be disabled otherwise | |
880 | default TRUE | |
881 | ||
882 | send_redirects - BOOLEAN | |
883 | Send redirects, if router. | |
884 | send_redirects for the interface will be enabled if at least one of | |
885 | conf/{all,interface}/send_redirects is set to TRUE, | |
886 | it will be disabled otherwise | |
887 | Default: TRUE | |
888 | ||
889 | bootp_relay - BOOLEAN | |
890 | Accept packets with source address 0.b.c.d destined | |
891 | not to this host as local ones. It is supposed, that | |
892 | BOOTP relay daemon will catch and forward such packets. | |
893 | conf/all/bootp_relay must also be set to TRUE to enable BOOTP relay | |
894 | for the interface | |
895 | default FALSE | |
896 | Not Implemented Yet. | |
897 | ||
898 | accept_source_route - BOOLEAN | |
899 | Accept packets with SRR option. | |
900 | conf/all/accept_source_route must also be set to TRUE to accept packets | |
901 | with SRR option on the interface | |
902 | default TRUE (router) | |
903 | FALSE (host) | |
904 | ||
8153a10c | 905 | accept_local - BOOLEAN |
c801e3cc DM |
906 | Accept packets with local source addresses. In combination |
907 | with suitable routing, this can be used to direct packets | |
908 | between two local interfaces over the wire and have them | |
909 | accepted properly. | |
910 | ||
911 | rp_filter must be set to a non-zero value in order for | |
912 | accept_local to have an effect. | |
913 | ||
8153a10c PM |
914 | default FALSE |
915 | ||
d0daebc3 TG |
916 | route_localnet - BOOLEAN |
917 | Do not consider loopback addresses as martian source or destination | |
918 | while routing. This enables the use of 127/8 for local routing purposes. | |
919 | default FALSE | |
920 | ||
c1cf8422 | 921 | rp_filter - INTEGER |
1da177e4 | 922 | 0 - No source validation. |
c1cf8422 SH |
923 | 1 - Strict mode as defined in RFC3704 Strict Reverse Path |
924 | Each incoming packet is tested against the FIB and if the interface | |
925 | is not the best reverse path the packet check will fail. | |
926 | By default failed packets are discarded. | |
927 | 2 - Loose mode as defined in RFC3704 Loose Reverse Path | |
928 | Each incoming packet's source address is also tested against the FIB | |
929 | and if the source address is not reachable via any interface | |
930 | the packet check will fail. | |
931 | ||
e18f5feb | 932 | Current recommended practice in RFC3704 is to enable strict mode |
bf869c30 | 933 | to prevent IP spoofing from DDos attacks. If using asymmetric routing |
e18f5feb | 934 | or other complicated routing, then loose mode is recommended. |
c1cf8422 | 935 | |
1f5865e7 SW |
936 | The max value from conf/{all,interface}/rp_filter is used |
937 | when doing source validation on the {interface}. | |
1da177e4 LT |
938 | |
939 | Default value is 0. Note that some distributions enable it | |
940 | in startup scripts. | |
941 | ||
942 | arp_filter - BOOLEAN | |
943 | 1 - Allows you to have multiple network interfaces on the same | |
944 | subnet, and have the ARPs for each interface be answered | |
945 | based on whether or not the kernel would route a packet from | |
946 | the ARP'd IP out that interface (therefore you must use source | |
947 | based routing for this to work). In other words it allows control | |
948 | of which cards (usually 1) will respond to an arp request. | |
949 | ||
950 | 0 - (default) The kernel can respond to arp requests with addresses | |
951 | from other interfaces. This may seem wrong but it usually makes | |
952 | sense, because it increases the chance of successful communication. | |
953 | IP addresses are owned by the complete host on Linux, not by | |
954 | particular interfaces. Only for more complex setups like load- | |
955 | balancing, does this behaviour cause problems. | |
956 | ||
957 | arp_filter for the interface will be enabled if at least one of | |
958 | conf/{all,interface}/arp_filter is set to TRUE, | |
959 | it will be disabled otherwise | |
960 | ||
961 | arp_announce - INTEGER | |
962 | Define different restriction levels for announcing the local | |
963 | source IP address from IP packets in ARP requests sent on | |
964 | interface: | |
965 | 0 - (default) Use any local address, configured on any interface | |
966 | 1 - Try to avoid local addresses that are not in the target's | |
967 | subnet for this interface. This mode is useful when target | |
968 | hosts reachable via this interface require the source IP | |
969 | address in ARP requests to be part of their logical network | |
970 | configured on the receiving interface. When we generate the | |
971 | request we will check all our subnets that include the | |
972 | target IP and will preserve the source address if it is from | |
973 | such subnet. If there is no such subnet we select source | |
974 | address according to the rules for level 2. | |
975 | 2 - Always use the best local address for this target. | |
976 | In this mode we ignore the source address in the IP packet | |
977 | and try to select local address that we prefer for talks with | |
978 | the target host. Such local address is selected by looking | |
979 | for primary IP addresses on all our subnets on the outgoing | |
980 | interface that include the target IP address. If no suitable | |
981 | local address is found we select the first local address | |
982 | we have on the outgoing interface or on all other interfaces, | |
983 | with the hope we will receive reply for our request and | |
984 | even sometimes no matter the source IP address we announce. | |
985 | ||
986 | The max value from conf/{all,interface}/arp_announce is used. | |
987 | ||
988 | Increasing the restriction level gives more chance for | |
989 | receiving answer from the resolved target while decreasing | |
990 | the level announces more valid sender's information. | |
991 | ||
992 | arp_ignore - INTEGER | |
993 | Define different modes for sending replies in response to | |
994 | received ARP requests that resolve local target IP addresses: | |
995 | 0 - (default): reply for any local target IP address, configured | |
996 | on any interface | |
997 | 1 - reply only if the target IP address is local address | |
998 | configured on the incoming interface | |
999 | 2 - reply only if the target IP address is local address | |
1000 | configured on the incoming interface and both with the | |
1001 | sender's IP address are part from same subnet on this interface | |
1002 | 3 - do not reply for local addresses configured with scope host, | |
1003 | only resolutions for global and link addresses are replied | |
1004 | 4-7 - reserved | |
1005 | 8 - do not reply for all local addresses | |
1006 | ||
1007 | The max value from conf/{all,interface}/arp_ignore is used | |
1008 | when ARP request is received on the {interface} | |
1009 | ||
eefef1cf SH |
1010 | arp_notify - BOOLEAN |
1011 | Define mode for notification of address and device changes. | |
1012 | 0 - (default): do nothing | |
3f8dc236 | 1013 | 1 - Generate gratuitous arp requests when device is brought up |
eefef1cf SH |
1014 | or hardware address changes. |
1015 | ||
c1b1bce8 | 1016 | arp_accept - BOOLEAN |
6d955180 OP |
1017 | Define behavior for gratuitous ARP frames who's IP is not |
1018 | already present in the ARP table: | |
1019 | 0 - don't create new entries in the ARP table | |
1020 | 1 - create new entries in the ARP table | |
1021 | ||
1022 | Both replies and requests type gratuitous arp will trigger the | |
1023 | ARP table to be updated, if this setting is on. | |
1024 | ||
1025 | If the ARP table already contains the IP address of the | |
1026 | gratuitous arp frame, the arp table will be updated regardless | |
1027 | if this setting is on or off. | |
1028 | ||
c1b1bce8 | 1029 | |
1da177e4 LT |
1030 | app_solicit - INTEGER |
1031 | The maximum number of probes to send to the user space ARP daemon | |
1032 | via netlink before dropping back to multicast probes (see | |
1033 | mcast_solicit). Defaults to 0. | |
1034 | ||
1035 | disable_policy - BOOLEAN | |
1036 | Disable IPSEC policy (SPD) for this interface | |
1037 | ||
1038 | disable_xfrm - BOOLEAN | |
1039 | Disable IPSEC encryption on this interface, whatever the policy | |
1040 | ||
1041 | ||
1042 | ||
1043 | tag - INTEGER | |
1044 | Allows you to write a number, which can be used as required. | |
1045 | Default value is 0. | |
1046 | ||
1da177e4 LT |
1047 | Alexey Kuznetsov. |
1048 | kuznet@ms2.inr.ac.ru | |
1049 | ||
1050 | Updated by: | |
1051 | Andi Kleen | |
1052 | ak@muc.de | |
1053 | Nicolas Delon | |
1054 | delon.nicolas@wanadoo.fr | |
1055 | ||
1056 | ||
1057 | ||
1058 | ||
1059 | /proc/sys/net/ipv6/* Variables: | |
1060 | ||
1061 | IPv6 has no global variables such as tcp_*. tcp_* settings under ipv4/ also | |
1062 | apply to IPv6 [XXX?]. | |
1063 | ||
1064 | bindv6only - BOOLEAN | |
1065 | Default value for IPV6_V6ONLY socket option, | |
e18f5feb | 1066 | which restricts use of the IPv6 socket to IPv6 communication |
1da177e4 LT |
1067 | only. |
1068 | TRUE: disable IPv4-mapped address feature | |
1069 | FALSE: enable IPv4-mapped address feature | |
1070 | ||
d5c073ca | 1071 | Default: FALSE (as specified in RFC3493) |
1da177e4 LT |
1072 | |
1073 | IPv6 Fragmentation: | |
1074 | ||
1075 | ip6frag_high_thresh - INTEGER | |
e18f5feb | 1076 | Maximum memory used to reassemble IPv6 fragments. When |
1da177e4 LT |
1077 | ip6frag_high_thresh bytes of memory is allocated for this purpose, |
1078 | the fragment handler will toss packets until ip6frag_low_thresh | |
1079 | is reached. | |
e18f5feb | 1080 | |
1da177e4 | 1081 | ip6frag_low_thresh - INTEGER |
e18f5feb | 1082 | See ip6frag_high_thresh |
1da177e4 LT |
1083 | |
1084 | ip6frag_time - INTEGER | |
1085 | Time in seconds to keep an IPv6 fragment in memory. | |
1086 | ||
1087 | ip6frag_secret_interval - INTEGER | |
e18f5feb | 1088 | Regeneration interval (in seconds) of the hash secret (or lifetime |
1da177e4 LT |
1089 | for the hash secret) for IPv6 fragments. |
1090 | Default: 600 | |
1091 | ||
1092 | conf/default/*: | |
1093 | Change the interface-specific default settings. | |
1094 | ||
1095 | ||
1096 | conf/all/*: | |
e18f5feb | 1097 | Change all the interface-specific settings. |
1da177e4 LT |
1098 | |
1099 | [XXX: Other special features than forwarding?] | |
1100 | ||
1101 | conf/all/forwarding - BOOLEAN | |
e18f5feb | 1102 | Enable global IPv6 forwarding between all interfaces. |
1da177e4 | 1103 | |
e18f5feb | 1104 | IPv4 and IPv6 work differently here; e.g. netfilter must be used |
1da177e4 LT |
1105 | to control which interfaces may forward packets and which not. |
1106 | ||
e18f5feb | 1107 | This also sets all interfaces' Host/Router setting |
1da177e4 LT |
1108 | 'forwarding' to the specified value. See below for details. |
1109 | ||
1110 | This referred to as global forwarding. | |
1111 | ||
fbea49e1 YH |
1112 | proxy_ndp - BOOLEAN |
1113 | Do proxy ndp. | |
1114 | ||
6fa3eb70 S |
1115 | fwmark_reflect - BOOLEAN |
1116 | Controls the fwmark of kernel-generated IPv6 reply packets that are not | |
1117 | associated with a socket for example, TCP RSTs or ICMPv6 echo replies). | |
1118 | If unset, these packets have a fwmark of zero. If set, they have the | |
1119 | fwmark of the packet they are replying to. Similarly affects the fwmark | |
1120 | used by internal routing lookups triggered by incoming packets, such as | |
1121 | the ones used for Path MTU Discovery. | |
1122 | Default: 0 | |
1123 | ||
1da177e4 LT |
1124 | conf/interface/*: |
1125 | Change special settings per interface. | |
1126 | ||
e18f5feb | 1127 | The functional behaviour for certain settings is different |
1da177e4 LT |
1128 | depending on whether local forwarding is enabled or not. |
1129 | ||
605b91c8 | 1130 | accept_ra - INTEGER |
1da177e4 | 1131 | Accept Router Advertisements; autoconfigure using them. |
e18f5feb | 1132 | |
026359bc TA |
1133 | It also determines whether or not to transmit Router |
1134 | Solicitations. If and only if the functional setting is to | |
1135 | accept Router Advertisements, Router Solicitations will be | |
1136 | transmitted. | |
1137 | ||
ae8abfa0 TG |
1138 | Possible values are: |
1139 | 0 Do not accept Router Advertisements. | |
1140 | 1 Accept Router Advertisements if forwarding is disabled. | |
1141 | 2 Overrule forwarding behaviour. Accept Router Advertisements | |
1142 | even if forwarding is enabled. | |
1143 | ||
1da177e4 LT |
1144 | Functional default: enabled if local forwarding is disabled. |
1145 | disabled if local forwarding is enabled. | |
1146 | ||
65f5c7c1 YH |
1147 | accept_ra_defrtr - BOOLEAN |
1148 | Learn default router in Router Advertisement. | |
1149 | ||
1150 | Functional default: enabled if accept_ra is enabled. | |
1151 | disabled if accept_ra is disabled. | |
1152 | ||
c4fd30eb | 1153 | accept_ra_pinfo - BOOLEAN |
2fe0ae78 | 1154 | Learn Prefix Information in Router Advertisement. |
c4fd30eb YH |
1155 | |
1156 | Functional default: enabled if accept_ra is enabled. | |
1157 | disabled if accept_ra is disabled. | |
1158 | ||
09c884d4 YH |
1159 | accept_ra_rt_info_max_plen - INTEGER |
1160 | Maximum prefix length of Route Information in RA. | |
1161 | ||
1162 | Route Information w/ prefix larger than or equal to this | |
1163 | variable shall be ignored. | |
1164 | ||
1165 | Functional default: 0 if accept_ra_rtr_pref is enabled. | |
1166 | -1 if accept_ra_rtr_pref is disabled. | |
1167 | ||
930d6ff2 YH |
1168 | accept_ra_rtr_pref - BOOLEAN |
1169 | Accept Router Preference in RA. | |
1170 | ||
1171 | Functional default: enabled if accept_ra is enabled. | |
1172 | disabled if accept_ra is disabled. | |
1173 | ||
1da177e4 LT |
1174 | accept_redirects - BOOLEAN |
1175 | Accept Redirects. | |
1176 | ||
1177 | Functional default: enabled if local forwarding is disabled. | |
1178 | disabled if local forwarding is enabled. | |
1179 | ||
0bcbc926 YH |
1180 | accept_source_route - INTEGER |
1181 | Accept source routing (routing extension header). | |
1182 | ||
bb4dbf9e | 1183 | >= 0: Accept only routing header type 2. |
0bcbc926 YH |
1184 | < 0: Do not accept routing header. |
1185 | ||
1186 | Default: 0 | |
1187 | ||
1da177e4 | 1188 | autoconf - BOOLEAN |
e18f5feb | 1189 | Autoconfigure addresses using Prefix Information in Router |
1da177e4 LT |
1190 | Advertisements. |
1191 | ||
c4fd30eb YH |
1192 | Functional default: enabled if accept_ra_pinfo is enabled. |
1193 | disabled if accept_ra_pinfo is disabled. | |
1da177e4 LT |
1194 | |
1195 | dad_transmits - INTEGER | |
1196 | The amount of Duplicate Address Detection probes to send. | |
1197 | Default: 1 | |
e18f5feb | 1198 | |
605b91c8 | 1199 | forwarding - INTEGER |
e18f5feb | 1200 | Configure interface-specific Host/Router behaviour. |
1da177e4 | 1201 | |
e18f5feb | 1202 | Note: It is recommended to have the same setting on all |
1da177e4 LT |
1203 | interfaces; mixed router/host scenarios are rather uncommon. |
1204 | ||
ae8abfa0 TG |
1205 | Possible values are: |
1206 | 0 Forwarding disabled | |
1207 | 1 Forwarding enabled | |
ae8abfa0 TG |
1208 | |
1209 | FALSE (0): | |
1da177e4 LT |
1210 | |
1211 | By default, Host behaviour is assumed. This means: | |
1212 | ||
1213 | 1. IsRouter flag is not set in Neighbour Advertisements. | |
026359bc TA |
1214 | 2. If accept_ra is TRUE (default), transmit Router |
1215 | Solicitations. | |
e18f5feb | 1216 | 3. If accept_ra is TRUE (default), accept Router |
1da177e4 LT |
1217 | Advertisements (and do autoconfiguration). |
1218 | 4. If accept_redirects is TRUE (default), accept Redirects. | |
1219 | ||
ae8abfa0 | 1220 | TRUE (1): |
1da177e4 | 1221 | |
e18f5feb | 1222 | If local forwarding is enabled, Router behaviour is assumed. |
1da177e4 LT |
1223 | This means exactly the reverse from the above: |
1224 | ||
1225 | 1. IsRouter flag is set in Neighbour Advertisements. | |
026359bc | 1226 | 2. Router Solicitations are not sent unless accept_ra is 2. |
ae8abfa0 | 1227 | 3. Router Advertisements are ignored unless accept_ra is 2. |
1da177e4 LT |
1228 | 4. Redirects are ignored. |
1229 | ||
ae8abfa0 TG |
1230 | Default: 0 (disabled) if global forwarding is disabled (default), |
1231 | otherwise 1 (enabled). | |
1da177e4 LT |
1232 | |
1233 | hop_limit - INTEGER | |
1234 | Default Hop Limit to set. | |
1235 | Default: 64 | |
1236 | ||
1237 | mtu - INTEGER | |
1238 | Default Maximum Transfer Unit | |
1239 | Default: 1280 (IPv6 required minimum) | |
1240 | ||
52e16356 YH |
1241 | router_probe_interval - INTEGER |
1242 | Minimum interval (in seconds) between Router Probing described | |
1243 | in RFC4191. | |
1244 | ||
1245 | Default: 60 | |
1246 | ||
1da177e4 LT |
1247 | router_solicitation_delay - INTEGER |
1248 | Number of seconds to wait after interface is brought up | |
1249 | before sending Router Solicitations. | |
1250 | Default: 1 | |
1251 | ||
1252 | router_solicitation_interval - INTEGER | |
1253 | Number of seconds to wait between Router Solicitations. | |
1254 | Default: 4 | |
1255 | ||
1256 | router_solicitations - INTEGER | |
e18f5feb | 1257 | Number of Router Solicitations to send until assuming no |
1da177e4 LT |
1258 | routers are present. |
1259 | Default: 3 | |
1260 | ||
1261 | use_tempaddr - INTEGER | |
1262 | Preference for Privacy Extensions (RFC3041). | |
1263 | <= 0 : disable Privacy Extensions | |
1264 | == 1 : enable Privacy Extensions, but prefer public | |
1265 | addresses over temporary addresses. | |
1266 | > 1 : enable Privacy Extensions and prefer temporary | |
1267 | addresses over public addresses. | |
1268 | Default: 0 (for most devices) | |
1269 | -1 (for point-to-point devices and loopback devices) | |
1270 | ||
1271 | temp_valid_lft - INTEGER | |
1272 | valid lifetime (in seconds) for temporary addresses. | |
1273 | Default: 604800 (7 days) | |
1274 | ||
1275 | temp_prefered_lft - INTEGER | |
1276 | Preferred lifetime (in seconds) for temporary addresses. | |
1277 | Default: 86400 (1 day) | |
1278 | ||
1279 | max_desync_factor - INTEGER | |
1280 | Maximum value for DESYNC_FACTOR, which is a random value | |
e18f5feb | 1281 | that ensures that clients don't synchronize with each |
1da177e4 LT |
1282 | other and generate new addresses at exactly the same time. |
1283 | value is in seconds. | |
1284 | Default: 600 | |
e18f5feb | 1285 | |
1da177e4 LT |
1286 | regen_max_retry - INTEGER |
1287 | Number of attempts before give up attempting to generate | |
1288 | valid temporary addresses. | |
1289 | Default: 5 | |
1290 | ||
1291 | max_addresses - INTEGER | |
e79dc484 BH |
1292 | Maximum number of autoconfigured addresses per interface. Setting |
1293 | to zero disables the limitation. It is not recommended to set this | |
1294 | value too large (or to zero) because it would be an easy way to | |
1295 | crash the kernel by allowing too many addresses to be created. | |
1da177e4 LT |
1296 | Default: 16 |
1297 | ||
778d80be | 1298 | disable_ipv6 - BOOLEAN |
9bdd8d40 BH |
1299 | Disable IPv6 operation. If accept_dad is set to 2, this value |
1300 | will be dynamically set to TRUE if DAD fails for the link-local | |
1301 | address. | |
778d80be YH |
1302 | Default: FALSE (enable IPv6 operation) |
1303 | ||
56d417b1 BH |
1304 | When this value is changed from 1 to 0 (IPv6 is being enabled), |
1305 | it will dynamically create a link-local address on the given | |
1306 | interface and start Duplicate Address Detection, if necessary. | |
1307 | ||
1308 | When this value is changed from 0 to 1 (IPv6 is being disabled), | |
1309 | it will dynamically delete all address on the given interface. | |
1310 | ||
1b34be74 YH |
1311 | accept_dad - INTEGER |
1312 | Whether to accept DAD (Duplicate Address Detection). | |
1313 | 0: Disable DAD | |
1314 | 1: Enable DAD (default) | |
1315 | 2: Enable DAD, and disable IPv6 operation if MAC-based duplicate | |
1316 | link-local address has been found. | |
1317 | ||
f7734fdf OP |
1318 | force_tllao - BOOLEAN |
1319 | Enable sending the target link-layer address option even when | |
1320 | responding to a unicast neighbor solicitation. | |
1321 | Default: FALSE | |
1322 | ||
1323 | Quoting from RFC 2461, section 4.4, Target link-layer address: | |
1324 | ||
1325 | "The option MUST be included for multicast solicitations in order to | |
1326 | avoid infinite Neighbor Solicitation "recursion" when the peer node | |
1327 | does not have a cache entry to return a Neighbor Advertisements | |
1328 | message. When responding to unicast solicitations, the option can be | |
1329 | omitted since the sender of the solicitation has the correct link- | |
1330 | layer address; otherwise it would not have be able to send the unicast | |
1331 | solicitation in the first place. However, including the link-layer | |
1332 | address in this case adds little overhead and eliminates a potential | |
1333 | race condition where the sender deletes the cached link-layer address | |
1334 | prior to receiving a response to a previous solicitation." | |
1335 | ||
db2b620a HFS |
1336 | ndisc_notify - BOOLEAN |
1337 | Define mode for notification of address and device changes. | |
1338 | 0 - (default): do nothing | |
1339 | 1 - Generate unsolicited neighbour advertisements when device is brought | |
1340 | up or hardware address changes. | |
1341 | ||
1da177e4 LT |
1342 | icmp/*: |
1343 | ratelimit - INTEGER | |
1344 | Limit the maximal rates for sending ICMPv6 packets. | |
6dbf4bca SH |
1345 | 0 to disable any limiting, |
1346 | otherwise the minimal space between responses in milliseconds. | |
1347 | Default: 1000 | |
1da177e4 LT |
1348 | |
1349 | ||
1350 | IPv6 Update by: | |
1351 | Pekka Savola <pekkas@netcore.fi> | |
1352 | YOSHIFUJI Hideaki / USAGI Project <yoshfuji@linux-ipv6.org> | |
1353 | ||
1354 | ||
1355 | /proc/sys/net/bridge/* Variables: | |
1356 | ||
1357 | bridge-nf-call-arptables - BOOLEAN | |
1358 | 1 : pass bridged ARP traffic to arptables' FORWARD chain. | |
1359 | 0 : disable this. | |
1360 | Default: 1 | |
1361 | ||
1362 | bridge-nf-call-iptables - BOOLEAN | |
1363 | 1 : pass bridged IPv4 traffic to iptables' chains. | |
1364 | 0 : disable this. | |
1365 | Default: 1 | |
1366 | ||
1367 | bridge-nf-call-ip6tables - BOOLEAN | |
1368 | 1 : pass bridged IPv6 traffic to ip6tables' chains. | |
1369 | 0 : disable this. | |
1370 | Default: 1 | |
1371 | ||
1372 | bridge-nf-filter-vlan-tagged - BOOLEAN | |
516299d2 MM |
1373 | 1 : pass bridged vlan-tagged ARP/IP/IPv6 traffic to {arp,ip,ip6}tables. |
1374 | 0 : disable this. | |
4981682c | 1375 | Default: 0 |
516299d2 MM |
1376 | |
1377 | bridge-nf-filter-pppoe-tagged - BOOLEAN | |
1378 | 1 : pass bridged pppoe-tagged IP/IPv6 traffic to {ip,ip6}tables. | |
1da177e4 | 1379 | 0 : disable this. |
4981682c | 1380 | Default: 0 |
1da177e4 | 1381 | |
4981682c PNA |
1382 | bridge-nf-pass-vlan-input-dev - BOOLEAN |
1383 | 1: if bridge-nf-filter-vlan-tagged is enabled, try to find a vlan | |
1384 | interface on the bridge and set the netfilter input device to the vlan. | |
1385 | This allows use of e.g. "iptables -i br0.1" and makes the REDIRECT | |
1386 | target work with vlan-on-top-of-bridge interfaces. When no matching | |
1387 | vlan interface is found, or this switch is off, the input device is | |
1388 | set to the bridge interface. | |
1389 | 0: disable bridge netfilter vlan interface lookup. | |
1390 | Default: 0 | |
1da177e4 | 1391 | |
32e8d494 VY |
1392 | proc/sys/net/sctp/* Variables: |
1393 | ||
1394 | addip_enable - BOOLEAN | |
1395 | Enable or disable extension of Dynamic Address Reconfiguration | |
1396 | (ADD-IP) functionality specified in RFC5061. This extension provides | |
1397 | the ability to dynamically add and remove new addresses for the SCTP | |
1398 | associations. | |
1399 | ||
1400 | 1: Enable extension. | |
1401 | ||
1402 | 0: Disable extension. | |
1403 | ||
1404 | Default: 0 | |
1405 | ||
1406 | addip_noauth_enable - BOOLEAN | |
1407 | Dynamic Address Reconfiguration (ADD-IP) requires the use of | |
1408 | authentication to protect the operations of adding or removing new | |
1409 | addresses. This requirement is mandated so that unauthorized hosts | |
1410 | would not be able to hijack associations. However, older | |
1411 | implementations may not have implemented this requirement while | |
1412 | allowing the ADD-IP extension. For reasons of interoperability, | |
1413 | we provide this variable to control the enforcement of the | |
1414 | authentication requirement. | |
1415 | ||
1416 | 1: Allow ADD-IP extension to be used without authentication. This | |
1417 | should only be set in a closed environment for interoperability | |
1418 | with older implementations. | |
1419 | ||
1420 | 0: Enforce the authentication requirement | |
1421 | ||
1422 | Default: 0 | |
1423 | ||
1424 | auth_enable - BOOLEAN | |
1425 | Enable or disable Authenticated Chunks extension. This extension | |
1426 | provides the ability to send and receive authenticated chunks and is | |
1427 | required for secure operation of Dynamic Address Reconfiguration | |
1428 | (ADD-IP) extension. | |
1429 | ||
1430 | 1: Enable this extension. | |
1431 | 0: Disable this extension. | |
1432 | ||
1433 | Default: 0 | |
1434 | ||
1435 | prsctp_enable - BOOLEAN | |
1436 | Enable or disable the Partial Reliability extension (RFC3758) which | |
1437 | is used to notify peers that a given DATA should no longer be expected. | |
1438 | ||
1439 | 1: Enable extension | |
1440 | 0: Disable | |
1441 | ||
1442 | Default: 1 | |
1443 | ||
1444 | max_burst - INTEGER | |
1445 | The limit of the number of new packets that can be initially sent. It | |
1446 | controls how bursty the generated traffic can be. | |
1447 | ||
1448 | Default: 4 | |
1449 | ||
1450 | association_max_retrans - INTEGER | |
1451 | Set the maximum number for retransmissions that an association can | |
1452 | attempt deciding that the remote end is unreachable. If this value | |
1453 | is exceeded, the association is terminated. | |
1454 | ||
1455 | Default: 10 | |
1456 | ||
1457 | max_init_retransmits - INTEGER | |
1458 | The maximum number of retransmissions of INIT and COOKIE-ECHO chunks | |
1459 | that an association will attempt before declaring the destination | |
1460 | unreachable and terminating. | |
1461 | ||
1462 | Default: 8 | |
1463 | ||
1464 | path_max_retrans - INTEGER | |
1465 | The maximum number of retransmissions that will be attempted on a given | |
1466 | path. Once this threshold is exceeded, the path is considered | |
1467 | unreachable, and new traffic will use a different path when the | |
1468 | association is multihomed. | |
1469 | ||
1470 | Default: 5 | |
1471 | ||
5aa93bcf NH |
1472 | pf_retrans - INTEGER |
1473 | The number of retransmissions that will be attempted on a given path | |
1474 | before traffic is redirected to an alternate transport (should one | |
1475 | exist). Note this is distinct from path_max_retrans, as a path that | |
1476 | passes the pf_retrans threshold can still be used. Its only | |
1477 | deprioritized when a transmission path is selected by the stack. This | |
1478 | setting is primarily used to enable fast failover mechanisms without | |
1479 | having to reduce path_max_retrans to a very low value. See: | |
1480 | http://www.ietf.org/id/draft-nishida-tsvwg-sctp-failover-05.txt | |
1481 | for details. Note also that a value of pf_retrans > path_max_retrans | |
1482 | disables this feature | |
1483 | ||
1484 | Default: 0 | |
1485 | ||
32e8d494 VY |
1486 | rto_initial - INTEGER |
1487 | The initial round trip timeout value in milliseconds that will be used | |
1488 | in calculating round trip times. This is the initial time interval | |
1489 | for retransmissions. | |
1490 | ||
1491 | Default: 3000 | |
1da177e4 | 1492 | |
32e8d494 VY |
1493 | rto_max - INTEGER |
1494 | The maximum value (in milliseconds) of the round trip timeout. This | |
1495 | is the largest time interval that can elapse between retransmissions. | |
1496 | ||
1497 | Default: 60000 | |
1498 | ||
1499 | rto_min - INTEGER | |
1500 | The minimum value (in milliseconds) of the round trip timeout. This | |
1501 | is the smallest time interval the can elapse between retransmissions. | |
1502 | ||
1503 | Default: 1000 | |
1504 | ||
1505 | hb_interval - INTEGER | |
1506 | The interval (in milliseconds) between HEARTBEAT chunks. These chunks | |
1507 | are sent at the specified interval on idle paths to probe the state of | |
1508 | a given path between 2 associations. | |
1509 | ||
1510 | Default: 30000 | |
1511 | ||
1512 | sack_timeout - INTEGER | |
1513 | The amount of time (in milliseconds) that the implementation will wait | |
1514 | to send a SACK. | |
1515 | ||
1516 | Default: 200 | |
1517 | ||
1518 | valid_cookie_life - INTEGER | |
1519 | The default lifetime of the SCTP cookie (in milliseconds). The cookie | |
1520 | is used during association establishment. | |
1521 | ||
1522 | Default: 60000 | |
1523 | ||
1524 | cookie_preserve_enable - BOOLEAN | |
1525 | Enable or disable the ability to extend the lifetime of the SCTP cookie | |
1526 | that is used during the establishment phase of SCTP association | |
1527 | ||
1528 | 1: Enable cookie lifetime extension. | |
1529 | 0: Disable | |
1530 | ||
1531 | Default: 1 | |
1532 | ||
3c68198e NH |
1533 | cookie_hmac_alg - STRING |
1534 | Select the hmac algorithm used when generating the cookie value sent by | |
1535 | a listening sctp socket to a connecting client in the INIT-ACK chunk. | |
1536 | Valid values are: | |
1537 | * md5 | |
1538 | * sha1 | |
1539 | * none | |
1540 | Ability to assign md5 or sha1 as the selected alg is predicated on the | |
3b09adcb | 1541 | configuration of those algorithms at build time (CONFIG_CRYPTO_MD5 and |
3c68198e NH |
1542 | CONFIG_CRYPTO_SHA1). |
1543 | ||
1544 | Default: Dependent on configuration. MD5 if available, else SHA1 if | |
1545 | available, else none. | |
1546 | ||
32e8d494 VY |
1547 | rcvbuf_policy - INTEGER |
1548 | Determines if the receive buffer is attributed to the socket or to | |
1549 | association. SCTP supports the capability to create multiple | |
1550 | associations on a single socket. When using this capability, it is | |
1551 | possible that a single stalled association that's buffering a lot | |
1552 | of data may block other associations from delivering their data by | |
1553 | consuming all of the receive buffer space. To work around this, | |
1554 | the rcvbuf_policy could be set to attribute the receiver buffer space | |
1555 | to each association instead of the socket. This prevents the described | |
1556 | blocking. | |
1557 | ||
1558 | 1: rcvbuf space is per association | |
3b09adcb | 1559 | 0: rcvbuf space is per socket |
32e8d494 VY |
1560 | |
1561 | Default: 0 | |
1562 | ||
1563 | sndbuf_policy - INTEGER | |
1564 | Similar to rcvbuf_policy above, this applies to send buffer space. | |
1565 | ||
1566 | 1: Send buffer is tracked per association | |
1567 | 0: Send buffer is tracked per socket. | |
1568 | ||
1569 | Default: 0 | |
1570 | ||
1571 | sctp_mem - vector of 3 INTEGERs: min, pressure, max | |
1572 | Number of pages allowed for queueing by all SCTP sockets. | |
1573 | ||
1574 | min: Below this number of pages SCTP is not bothered about its | |
1575 | memory appetite. When amount of memory allocated by SCTP exceeds | |
1576 | this number, SCTP starts to moderate memory usage. | |
1577 | ||
1578 | pressure: This value was introduced to follow format of tcp_mem. | |
1579 | ||
1580 | max: Number of pages allowed for queueing by all SCTP sockets. | |
1581 | ||
1582 | Default is calculated at boot time from amount of available memory. | |
1583 | ||
1584 | sctp_rmem - vector of 3 INTEGERs: min, default, max | |
a6e1204b MM |
1585 | Only the first value ("min") is used, "default" and "max" are |
1586 | ignored. | |
1587 | ||
1588 | min: Minimal size of receive buffer used by SCTP socket. | |
1589 | It is guaranteed to each SCTP socket (but not association) even | |
1590 | under moderate memory pressure. | |
1591 | ||
1592 | Default: 1 page | |
32e8d494 VY |
1593 | |
1594 | sctp_wmem - vector of 3 INTEGERs: min, default, max | |
a6e1204b | 1595 | Currently this tunable has no effect. |
32e8d494 | 1596 | |
72388433 BD |
1597 | addr_scope_policy - INTEGER |
1598 | Control IPv4 address scoping - draft-stewart-tsvwg-sctp-ipv4-00 | |
1599 | ||
1600 | 0 - Disable IPv4 address scoping | |
1601 | 1 - Enable IPv4 address scoping | |
1602 | 2 - Follow draft but allow IPv4 private addresses | |
1603 | 3 - Follow draft but allow IPv4 link local addresses | |
1604 | ||
1605 | Default: 1 | |
1606 | ||
1da177e4 | 1607 | |
4edc2f34 | 1608 | /proc/sys/net/core/* |
c60f6aa8 | 1609 | Please see: Documentation/sysctl/net.txt for descriptions of these entries. |
705efc3b | 1610 | |
4edc2f34 SH |
1611 | |
1612 | /proc/sys/net/unix/* | |
705efc3b WT |
1613 | max_dgram_qlen - INTEGER |
1614 | The maximum length of dgram socket receive queue | |
1615 | ||
1616 | Default: 10 | |
1617 | ||
1618 | ||
1619 | UNDOCUMENTED: | |
4edc2f34 SH |
1620 | |
1621 | /proc/sys/net/irda/* | |
1622 | fast_poll_increase FIXME | |
1623 | warn_noreply_time FIXME | |
1624 | discovery_slots FIXME | |
1625 | slot_timeout FIXME | |
1626 | max_baud_rate FIXME | |
1627 | discovery_timeout FIXME | |
1628 | lap_keepalive_time FIXME | |
1629 | max_noreply_time FIXME | |
1630 | max_tx_data_size FIXME | |
1631 | max_tx_window FIXME | |
1632 | min_tx_turn_time FIXME |