1.. SPDX-License-Identifier: GPL-2.0 2 3========= 4IP Sysctl 5========= 6 7/proc/sys/net/ipv4/* Variables 8============================== 9 10ip_forward - BOOLEAN 11 Forward Packets between interfaces. 12 13 This variable is special, its change resets all configuration 14 parameters to their default state (RFC1122 for hosts, RFC1812 15 for routers) 16 17 Possible values: 18 19 - 0 (disabled) 20 - 1 (enabled) 21 22 Default: 0 (disabled) 23 24ip_default_ttl - INTEGER 25 Default value of TTL field (Time To Live) for outgoing (but not 26 forwarded) IP packets. Should be between 1 and 255 inclusive. 27 Default: 64 (as recommended by RFC1700) 28 29ip_no_pmtu_disc - INTEGER 30 Disable Path MTU Discovery. If enabled in mode 1 and a 31 fragmentation-required ICMP is received, the PMTU to this 32 destination will be set to the smallest of the old MTU to 33 this destination and min_pmtu (see below). You will need 34 to raise min_pmtu to the smallest interface MTU on your system 35 manually if you want to avoid locally generated fragments. 36 37 In mode 2 incoming Path MTU Discovery messages will be 38 discarded. Outgoing frames are handled the same as in mode 1, 39 implicitly setting IP_PMTUDISC_DONT on every created socket. 40 41 Mode 3 is a hardened pmtu discover mode. The kernel will only 42 accept fragmentation-needed errors if the underlying protocol 43 can verify them besides a plain socket lookup. Current 44 protocols for which pmtu events will be honored are TCP and 45 SCTP as they verify e.g. the sequence number or the 46 association. This mode should not be enabled globally but is 47 only intended to secure e.g. name servers in namespaces where 48 TCP path mtu must still work but path MTU information of other 49 protocols should be discarded. If enabled globally this mode 50 could break other protocols. 51 52 Possible values: 0-3 53 54 Default: FALSE 55 56min_pmtu - INTEGER 57 default 552 - minimum Path MTU. Unless this is changed manually, 58 each cached pmtu will never be lower than this setting. 59 60ip_forward_use_pmtu - BOOLEAN 61 By default we don't trust protocol path MTUs while forwarding 62 because they could be easily forged and can lead to unwanted 63 fragmentation by the router. 64 You only need to enable this if you have user-space software 65 which tries to discover path mtus by itself and depends on the 66 kernel honoring this information. This is normally not the 67 case. 68 69 Possible values: 70 71 - 0 (disabled) 72 - 1 (enabled) 73 74 Default: 0 (disabled) 75 76fwmark_reflect - BOOLEAN 77 Controls the fwmark of kernel-generated IPv4 reply packets that are not 78 associated with a socket for example, TCP RSTs or ICMP echo replies). 79 If disabled, these packets have a fwmark of zero. If enabled, they have the 80 fwmark of the packet they are replying to. 81 82 Possible values: 83 84 - 0 (disabled) 85 - 1 (enabled) 86 87 Default: 0 (disabled) 88 89fib_multipath_use_neigh - BOOLEAN 90 Use status of existing neighbor entry when determining nexthop for 91 multipath routes. If disabled, neighbor information is not used and 92 packets could be directed to a failed nexthop. Only valid for kernels 93 built with CONFIG_IP_ROUTE_MULTIPATH enabled. 94 95 Possible values: 96 97 - 0 (disabled) 98 - 1 (enabled) 99 100 Default: 0 (disabled) 101 102fib_multipath_hash_policy - INTEGER 103 Controls which hash policy to use for multipath routes. Only valid 104 for kernels built with CONFIG_IP_ROUTE_MULTIPATH enabled. 105 106 Default: 0 (Layer 3) 107 108 Possible values: 109 110 - 0 - Layer 3 111 - 1 - Layer 4 112 - 2 - Layer 3 or inner Layer 3 if present 113 - 3 - Custom multipath hash. Fields used for multipath hash calculation 114 are determined by fib_multipath_hash_fields sysctl 115 116fib_multipath_hash_fields - UNSIGNED INTEGER 117 When fib_multipath_hash_policy is set to 3 (custom multipath hash), the 118 fields used for multipath hash calculation are determined by this 119 sysctl. 120 121 This value is a bitmask which enables various fields for multipath hash 122 calculation. 123 124 Possible fields are: 125 126 ====== ============================ 127 0x0001 Source IP address 128 0x0002 Destination IP address 129 0x0004 IP protocol 130 0x0008 Unused (Flow Label) 131 0x0010 Source port 132 0x0020 Destination port 133 0x0040 Inner source IP address 134 0x0080 Inner destination IP address 135 0x0100 Inner IP protocol 136 0x0200 Inner Flow Label 137 0x0400 Inner source port 138 0x0800 Inner destination port 139 ====== ============================ 140 141 Default: 0x0007 (source IP, destination IP and IP protocol) 142 143fib_multipath_hash_seed - UNSIGNED INTEGER 144 The seed value used when calculating hash for multipath routes. Applies 145 to both IPv4 and IPv6 datapath. Only present for kernels built with 146 CONFIG_IP_ROUTE_MULTIPATH enabled. 147 148 When set to 0, the seed value used for multipath routing defaults to an 149 internal random-generated one. 150 151 The actual hashing algorithm is not specified -- there is no guarantee 152 that a next hop distribution effected by a given seed will keep stable 153 across kernel versions. 154 155 Default: 0 (random) 156 157fib_sync_mem - UNSIGNED INTEGER 158 Amount of dirty memory from fib entries that can be backlogged before 159 synchronize_rcu is forced. 160 161 Default: 512kB Minimum: 64kB Maximum: 64MB 162 163ip_forward_update_priority - INTEGER 164 Whether to update SKB priority from "TOS" field in IPv4 header after it 165 is forwarded. The new SKB priority is mapped from TOS field value 166 according to an rt_tos2priority table (see e.g. man tc-prio). 167 168 Default: 1 (Update priority.) 169 170 Possible values: 171 172 - 0 - Do not update priority. 173 - 1 - Update priority. 174 175route/max_size - INTEGER 176 Maximum number of routes allowed in the kernel. Increase 177 this when using large numbers of interfaces and/or routes. 178 179 From linux kernel 3.6 onwards, this is deprecated for ipv4 180 as route cache is no longer used. 181 182 From linux kernel 6.3 onwards, this is deprecated for ipv6 183 as garbage collection manages cached route entries. 184 185neigh/default/gc_thresh1 - INTEGER 186 Minimum number of entries to keep. Garbage collector will not 187 purge entries if there are fewer than this number. 188 189 Default: 128 190 191neigh/default/gc_thresh2 - INTEGER 192 Threshold when garbage collector becomes more aggressive about 193 purging entries. Entries older than 5 seconds will be cleared 194 when over this number. 195 196 Default: 512 197 198neigh/default/gc_thresh3 - INTEGER 199 Maximum number of non-PERMANENT neighbor entries allowed. Increase 200 this when using large numbers of interfaces and when communicating 201 with large numbers of directly-connected peers. 202 203 Default: 1024 204 205neigh/default/unres_qlen_bytes - INTEGER 206 The maximum number of bytes which may be used by packets 207 queued for each unresolved address by other network layers. 208 (added in linux 3.3) 209 210 Setting negative value is meaningless and will return error. 211 212 Default: SK_WMEM_DEFAULT, (same as net.core.wmem_default). 213 214 Exact value depends on architecture and kernel options, 215 but should be enough to allow queuing 256 packets 216 of medium size. 217 218neigh/default/unres_qlen - INTEGER 219 The maximum number of packets which may be queued for each 220 unresolved address by other network layers. 221 222 (deprecated in linux 3.3) : use unres_qlen_bytes instead. 223 224 Prior to linux 3.3, the default value is 3 which may cause 225 unexpected packet loss. The current default value is calculated 226 according to default value of unres_qlen_bytes and true size of 227 packet. 228 229 Default: 101 230 231neigh/default/interval_probe_time_ms - INTEGER 232 The probe interval for neighbor entries with NTF_MANAGED flag, 233 the min value is 1. 234 235 Default: 5000 236 237mtu_expires - INTEGER 238 Time, in seconds, that cached PMTU information is kept. 239 240min_adv_mss - INTEGER 241 The advertised MSS depends on the first hop route MTU, but will 242 never be lower than this setting. 243 244fib_notify_on_flag_change - INTEGER 245 Whether to emit RTM_NEWROUTE notifications whenever RTM_F_OFFLOAD/ 246 RTM_F_TRAP/RTM_F_OFFLOAD_FAILED flags are changed. 247 248 After installing a route to the kernel, user space receives an 249 acknowledgment, which means the route was installed in the kernel, 250 but not necessarily in hardware. 251 It is also possible for a route already installed in hardware to change 252 its action and therefore its flags. For example, a host route that is 253 trapping packets can be "promoted" to perform decapsulation following 254 the installation of an IPinIP/VXLAN tunnel. 255 The notifications will indicate to user-space the state of the route. 256 257 Default: 0 (Do not emit notifications.) 258 259 Possible values: 260 261 - 0 - Do not emit notifications. 262 - 1 - Emit notifications. 263 - 2 - Emit notifications only for RTM_F_OFFLOAD_FAILED flag change. 264 265IP Fragmentation: 266 267ipfrag_high_thresh - LONG INTEGER 268 Maximum memory used to reassemble IP fragments. 269 270ipfrag_low_thresh - LONG INTEGER 271 (Obsolete since linux-4.17) 272 Maximum memory used to reassemble IP fragments before the kernel 273 begins to remove incomplete fragment queues to free up resources. 274 The kernel still accepts new fragments for defragmentation. 275 276ipfrag_time - INTEGER 277 Time in seconds to keep an IP fragment in memory. 278 279ipfrag_max_dist - INTEGER 280 ipfrag_max_dist is a non-negative integer value which defines the 281 maximum "disorder" which is allowed among fragments which share a 282 common IP source address. Note that reordering of packets is 283 not unusual, but if a large number of fragments arrive from a source 284 IP address while a particular fragment queue remains incomplete, it 285 probably indicates that one or more fragments belonging to that queue 286 have been lost. When ipfrag_max_dist is positive, an additional check 287 is done on fragments before they are added to a reassembly queue - if 288 ipfrag_max_dist (or more) fragments have arrived from a particular IP 289 address between additions to any IP fragment queue using that source 290 address, it's presumed that one or more fragments in the queue are 291 lost. The existing fragment queue will be dropped, and a new one 292 started. An ipfrag_max_dist value of zero disables this check. 293 294 Using a very small value, e.g. 1 or 2, for ipfrag_max_dist can 295 result in unnecessarily dropping fragment queues when normal 296 reordering of packets occurs, which could lead to poor application 297 performance. Using a very large value, e.g. 50000, increases the 298 likelihood of incorrectly reassembling IP fragments that originate 299 from different IP datagrams, which could result in data corruption. 300 Default: 64 301 302bc_forwarding - INTEGER 303 bc_forwarding enables the feature described in rfc1812#section-5.3.5.2 304 and rfc2644. It allows the router to forward directed broadcast. 305 To enable this feature, the 'all' entry and the input interface entry 306 should be set to 1. 307 Default: 0 308 309INET peer storage 310================= 311 312inet_peer_threshold - INTEGER 313 The approximate size of the storage. Starting from this threshold 314 entries will be thrown aggressively. This threshold also determines 315 entries' time-to-live and time intervals between garbage collection 316 passes. More entries, less time-to-live, less GC interval. 317 318inet_peer_minttl - INTEGER 319 Minimum time-to-live of entries. Should be enough to cover fragment 320 time-to-live on the reassembling side. This minimum time-to-live is 321 guaranteed if the pool size is less than inet_peer_threshold. 322 Measured in seconds. 323 324inet_peer_maxttl - INTEGER 325 Maximum time-to-live of entries. Unused entries will expire after 326 this period of time if there is no memory pressure on the pool (i.e. 327 when the number of entries in the pool is very small). 328 Measured in seconds. 329 330TCP variables 331============= 332 333somaxconn - INTEGER 334 Limit of socket listen() backlog, known in userspace as SOMAXCONN. 335 Defaults to 4096. (Was 128 before linux-5.4) 336 See also tcp_max_syn_backlog for additional tuning for TCP sockets. 337 338tcp_abort_on_overflow - BOOLEAN 339 If listening service is too slow to accept new connections, 340 reset them. Default state is FALSE. It means that if overflow 341 occurred due to a burst, connection will recover. Enable this 342 option _only_ if you are really sure that listening daemon 343 cannot be tuned to accept connections faster. Enabling this 344 option can harm clients of your server. 345 346tcp_adv_win_scale - INTEGER 347 Obsolete since linux-6.6 348 Count buffering overhead as bytes/2^tcp_adv_win_scale 349 (if tcp_adv_win_scale > 0) or bytes-bytes/2^(-tcp_adv_win_scale), 350 if it is <= 0. 351 352 Possible values are [-31, 31], inclusive. 353 354 Default: 1 355 356tcp_allowed_congestion_control - STRING 357 Show/set the congestion control choices available to non-privileged 358 processes. The list is a subset of those listed in 359 tcp_available_congestion_control. 360 361 Default is "reno" and the default setting (tcp_congestion_control). 362 363tcp_app_win - INTEGER 364 Reserve max(window/2^tcp_app_win, mss) of window for application 365 buffer. Value 0 is special, it means that nothing is reserved. 366 367 Possible values are [0, 31], inclusive. 368 369 Default: 31 370 371tcp_autocorking - BOOLEAN 372 Enable TCP auto corking : 373 When applications do consecutive small write()/sendmsg() system calls, 374 we try to coalesce these small writes as much as possible, to lower 375 total amount of sent packets. This is done if at least one prior 376 packet for the flow is waiting in Qdisc queues or device transmit 377 queue. Applications can still use TCP_CORK for optimal behavior 378 when they know how/when to uncork their sockets. 379 380 Possible values: 381 382 - 0 (disabled) 383 - 1 (enabled) 384 385 Default: 1 (enabled) 386 387tcp_available_congestion_control - STRING 388 Shows the available congestion control choices that are registered. 389 More congestion control algorithms may be available as modules, 390 but not loaded. 391 392tcp_base_mss - INTEGER 393 The initial value of search_low to be used by the packetization layer 394 Path MTU discovery (MTU probing). If MTU probing is enabled, 395 this is the initial MSS used by the connection. 396 397tcp_mtu_probe_floor - INTEGER 398 If MTU probing is enabled this caps the minimum MSS used for search_low 399 for the connection. 400 401 Default : 48 402 403tcp_min_snd_mss - INTEGER 404 TCP SYN and SYNACK messages usually advertise an ADVMSS option, 405 as described in RFC 1122 and RFC 6691. 406 407 If this ADVMSS option is smaller than tcp_min_snd_mss, 408 it is silently capped to tcp_min_snd_mss. 409 410 Default : 48 (at least 8 bytes of payload per segment) 411 412tcp_congestion_control - STRING 413 Set the congestion control algorithm to be used for new 414 connections. The algorithm "reno" is always available, but 415 additional choices may be available based on kernel configuration. 416 Default is set as part of kernel configuration. 417 For passive connections, the listener congestion control choice 418 is inherited. 419 420 [see setsockopt(listenfd, SOL_TCP, TCP_CONGESTION, "name" ...) ] 421 422tcp_dsack - BOOLEAN 423 Allows TCP to send "duplicate" SACKs. 424 425 Possible values: 426 427 - 0 (disabled) 428 - 1 (enabled) 429 430 Default: 1 (enabled) 431 432tcp_early_retrans - INTEGER 433 Tail loss probe (TLP) converts RTOs occurring due to tail 434 losses into fast recovery (RFC8985). Note that 435 TLP requires RACK to function properly (see tcp_recovery below) 436 437 Possible values: 438 439 - 0 disables TLP 440 - 3 or 4 enables TLP 441 442 Default: 3 443 444tcp_ecn - INTEGER 445 Control use of Explicit Congestion Notification (ECN) by TCP. 446 ECN is used only when both ends of the TCP connection indicate support 447 for it. This feature is useful in avoiding losses due to congestion by 448 allowing supporting routers to signal congestion before having to drop 449 packets. A host that supports ECN both sends ECN at the IP layer and 450 feeds back ECN at the TCP layer. The highest variant of ECN feedback 451 that both peers support is chosen by the ECN negotiation (Accurate ECN, 452 ECN, or no ECN). 453 454 The highest negotiated variant for incoming connection requests 455 and the highest variant requested by outgoing connection 456 attempts: 457 458 ===== ==================== ==================== 459 Value Incoming connections Outgoing connections 460 ===== ==================== ==================== 461 0 No ECN No ECN 462 1 ECN ECN 463 2 ECN No ECN 464 3 AccECN AccECN 465 4 AccECN ECN 466 5 AccECN No ECN 467 ===== ==================== ==================== 468 469 Default: 2 470 471tcp_ecn_option - INTEGER 472 Control Accurate ECN (AccECN) option sending when AccECN has been 473 successfully negotiated during handshake. Send logic inhibits 474 sending AccECN options regarless of this setting when no AccECN 475 option has been seen for the reverse direction. 476 477 Possible values are: 478 479 = ============================================================ 480 0 Never send AccECN option. This also disables sending AccECN 481 option in SYN/ACK during handshake. 482 1 Send AccECN option sparingly according to the minimum option 483 rules outlined in draft-ietf-tcpm-accurate-ecn. 484 2 Send AccECN option on every packet whenever it fits into TCP 485 option space. 486 = ============================================================ 487 488 Default: 2 489 490tcp_ecn_option_beacon - INTEGER 491 Control Accurate ECN (AccECN) option sending frequency per RTT and it 492 takes effect only when tcp_ecn_option is set to 2. 493 494 Default: 3 (AccECN will be send at least 3 times per RTT) 495 496tcp_ecn_fallback - BOOLEAN 497 If the kernel detects that ECN connection misbehaves, enable fall 498 back to non-ECN. Currently, this knob implements the fallback 499 from RFC3168, section 6.1.1.1., but we reserve that in future, 500 additional detection mechanisms could be implemented under this 501 knob. The value is not used, if tcp_ecn or per route (or congestion 502 control) ECN settings are disabled. 503 504 Possible values: 505 506 - 0 (disabled) 507 - 1 (enabled) 508 509 Default: 1 (enabled) 510 511tcp_fack - BOOLEAN 512 This is a legacy option, it has no effect anymore. 513 514tcp_fin_timeout - INTEGER 515 The length of time an orphaned (no longer referenced by any 516 application) connection will remain in the FIN_WAIT_2 state 517 before it is aborted at the local end. While a perfectly 518 valid "receive only" state for an un-orphaned connection, an 519 orphaned connection in FIN_WAIT_2 state could otherwise wait 520 forever for the remote to close its end of the connection. 521 522 Cf. tcp_max_orphans 523 524 Default: 60 seconds 525 526tcp_frto - INTEGER 527 Enables Forward RTO-Recovery (F-RTO) defined in RFC5682. 528 F-RTO is an enhanced recovery algorithm for TCP retransmission 529 timeouts. It is particularly beneficial in networks where the 530 RTT fluctuates (e.g., wireless). F-RTO is sender-side only 531 modification. It does not require any support from the peer. 532 533 By default it's enabled with a non-zero value. 0 disables F-RTO. 534 535tcp_fwmark_accept - BOOLEAN 536 If enabled, incoming connections to listening sockets that do not have a 537 socket mark will set the mark of the accepting socket to the fwmark of 538 the incoming SYN packet. This will cause all packets on that connection 539 (starting from the first SYNACK) to be sent with that fwmark. The 540 listening socket's mark is unchanged. Listening sockets that already 541 have a fwmark set via setsockopt(SOL_SOCKET, SO_MARK, ...) are 542 unaffected. 543 544 Possible values: 545 546 - 0 (disabled) 547 - 1 (enabled) 548 549 Default: 0 (disabled) 550 551tcp_invalid_ratelimit - INTEGER 552 Limit the maximal rate for sending duplicate acknowledgments 553 in response to incoming TCP packets that are for an existing 554 connection but that are invalid due to any of these reasons: 555 556 (a) out-of-window sequence number, 557 (b) out-of-window acknowledgment number, or 558 (c) PAWS (Protection Against Wrapped Sequence numbers) check failure 559 560 This can help mitigate simple "ack loop" DoS attacks, wherein 561 a buggy or malicious middlebox or man-in-the-middle can 562 rewrite TCP header fields in manner that causes each endpoint 563 to think that the other is sending invalid TCP segments, thus 564 causing each side to send an unterminating stream of duplicate 565 acknowledgments for invalid segments. 566 567 Using 0 disables rate-limiting of dupacks in response to 568 invalid segments; otherwise this value specifies the minimal 569 space between sending such dupacks, in milliseconds. 570 571 Default: 500 (milliseconds). 572 573tcp_keepalive_time - INTEGER 574 How often TCP sends out keepalive messages when keepalive is enabled. 575 Default: 2hours. 576 577tcp_keepalive_probes - INTEGER 578 How many keepalive probes TCP sends out, until it decides that the 579 connection is broken. Default value: 9. 580 581tcp_keepalive_intvl - INTEGER 582 How frequently the probes are send out. Multiplied by 583 tcp_keepalive_probes it is time to kill not responding connection, 584 after probes started. Default value: 75sec i.e. connection 585 will be aborted after ~11 minutes of retries. 586 587tcp_l3mdev_accept - BOOLEAN 588 Enables child sockets to inherit the L3 master device index. 589 Enabling this option allows a "global" listen socket to work 590 across L3 master domains (e.g., VRFs) with connected sockets 591 derived from the listen socket to be bound to the L3 domain in 592 which the packets originated. Only valid when the kernel was 593 compiled with CONFIG_NET_L3_MASTER_DEV. 594 595 Possible values: 596 597 - 0 (disabled) 598 - 1 (enabled) 599 600 Default: 0 (disabled) 601 602tcp_low_latency - BOOLEAN 603 This is a legacy option, it has no effect anymore. 604 605tcp_max_orphans - INTEGER 606 Maximal number of TCP sockets not attached to any user file handle, 607 held by system. If this number is exceeded orphaned connections are 608 reset immediately and warning is printed. This limit exists 609 only to prevent simple DoS attacks, you _must_ not rely on this 610 or lower the limit artificially, but rather increase it 611 (probably, after increasing installed memory), 612 if network conditions require more than default value, 613 and tune network services to linger and kill such states 614 more aggressively. Let me to remind again: each orphan eats 615 up to ~64K of unswappable memory. 616 617tcp_max_syn_backlog - INTEGER 618 Maximal number of remembered connection requests (SYN_RECV), 619 which have not received an acknowledgment from connecting client. 620 621 This is a per-listener limit. 622 623 The minimal value is 128 for low memory machines, and it will 624 increase in proportion to the memory of machine. 625 626 If server suffers from overload, try increasing this number. 627 628 Remember to also check /proc/sys/net/core/somaxconn 629 A SYN_RECV request socket consumes about 304 bytes of memory. 630 631tcp_max_tw_buckets - INTEGER 632 Maximal number of timewait sockets held by system simultaneously. 633 If this number is exceeded time-wait socket is immediately destroyed 634 and warning is printed. This limit exists only to prevent 635 simple DoS attacks, you _must_ not lower the limit artificially, 636 but rather increase it (probably, after increasing installed memory), 637 if network conditions require more than default value. 638 639tcp_mem - vector of 3 INTEGERs: min, pressure, max 640 min: below this number of pages TCP is not bothered about its 641 memory appetite. 642 643 pressure: when amount of memory allocated by TCP exceeds this number 644 of pages, TCP moderates its memory consumption and enters memory 645 pressure mode, which is exited when memory consumption falls 646 under "min". 647 648 max: number of pages allowed for queueing by all TCP sockets. 649 650 Defaults are calculated at boot time from amount of available 651 memory. 652 653tcp_min_rtt_wlen - INTEGER 654 The window length of the windowed min filter to track the minimum RTT. 655 A shorter window lets a flow more quickly pick up new (higher) 656 minimum RTT when it is moved to a longer path (e.g., due to traffic 657 engineering). A longer window makes the filter more resistant to RTT 658 inflations such as transient congestion. The unit is seconds. 659 660 Possible values: 0 - 86400 (1 day) 661 662 Default: 300 663 664tcp_moderate_rcvbuf - BOOLEAN 665 If enabled, TCP performs receive buffer auto-tuning, attempting to 666 automatically size the buffer (no greater than tcp_rmem[2]) to 667 match the size required by the path for full throughput. 668 669 Possible values: 670 671 - 0 (disabled) 672 - 1 (enabled) 673 674 Default: 1 (enabled) 675 676tcp_mtu_probing - INTEGER 677 Controls TCP Packetization-Layer Path MTU Discovery. Takes three 678 values: 679 680 - 0 - Disabled 681 - 1 - Disabled by default, enabled when an ICMP black hole detected 682 - 2 - Always enabled, use initial MSS of tcp_base_mss. 683 684tcp_probe_interval - UNSIGNED INTEGER 685 Controls how often to start TCP Packetization-Layer Path MTU 686 Discovery reprobe. The default is reprobing every 10 minutes as 687 per RFC4821. 688 689tcp_probe_threshold - INTEGER 690 Controls when TCP Packetization-Layer Path MTU Discovery probing 691 will stop in respect to the width of search range in bytes. Default 692 is 8 bytes. 693 694tcp_no_metrics_save - BOOLEAN 695 By default, TCP saves various connection metrics in the route cache 696 when the connection closes, so that connections established in the 697 near future can use these to set initial conditions. Usually, this 698 increases overall performance, but may sometimes cause performance 699 degradation. If enabled, TCP will not cache metrics on closing 700 connections. 701 702 Possible values: 703 704 - 0 (disabled) 705 - 1 (enabled) 706 707 Default: 0 (disabled) 708 709tcp_no_ssthresh_metrics_save - BOOLEAN 710 Controls whether TCP saves ssthresh metrics in the route cache. 711 If enabled, ssthresh metrics are disabled. 712 713 Possible values: 714 715 - 0 (disabled) 716 - 1 (enabled) 717 718 Default: 1 (enabled) 719 720tcp_orphan_retries - INTEGER 721 This value influences the timeout of a locally closed TCP connection, 722 when RTO retransmissions remain unacknowledged. 723 See tcp_retries2 for more details. 724 725 The default value is 8. 726 727 If your machine is a loaded WEB server, 728 you should think about lowering this value, such sockets 729 may consume significant resources. Cf. tcp_max_orphans. 730 731tcp_recovery - INTEGER 732 This value is a bitmap to enable various experimental loss recovery 733 features. 734 735 ========= ============================================================= 736 RACK: 0x1 enables RACK loss detection, for fast detection of lost 737 retransmissions and tail drops, and resilience to 738 reordering. currently, setting this bit to 0 has no 739 effect, since RACK is the only supported loss detection 740 algorithm. 741 742 RACK: 0x2 makes RACK's reordering window static (min_rtt/4). 743 744 RACK: 0x4 disables RACK's DUPACK threshold heuristic 745 ========= ============================================================= 746 747 Default: 0x1 748 749tcp_reflect_tos - BOOLEAN 750 For listening sockets, reuse the DSCP value of the initial SYN message 751 for outgoing packets. This allows to have both directions of a TCP 752 stream to use the same DSCP value, assuming DSCP remains unchanged for 753 the lifetime of the connection. 754 755 This options affects both IPv4 and IPv6. 756 757 Possible values: 758 759 - 0 (disabled) 760 - 1 (enabled) 761 762 Default: 0 (disabled) 763 764tcp_reordering - INTEGER 765 Initial reordering level of packets in a TCP stream. 766 TCP stack can then dynamically adjust flow reordering level 767 between this initial value and tcp_max_reordering 768 769 Default: 3 770 771tcp_max_reordering - INTEGER 772 Maximal reordering level of packets in a TCP stream. 773 300 is a fairly conservative value, but you might increase it 774 if paths are using per packet load balancing (like bonding rr mode) 775 776 Default: 300 777 778tcp_retrans_collapse - BOOLEAN 779 Bug-to-bug compatibility with some broken printers. 780 On retransmit try to send bigger packets to work around bugs in 781 certain TCP stacks. 782 783 Possible values: 784 785 - 0 (disabled) 786 - 1 (enabled) 787 788 Default: 1 (enabled) 789 790tcp_retries1 - INTEGER 791 This value influences the time, after which TCP decides, that 792 something is wrong due to unacknowledged RTO retransmissions, 793 and reports this suspicion to the network layer. 794 See tcp_retries2 for more details. 795 796 RFC 1122 recommends at least 3 retransmissions, which is the 797 default. 798 799tcp_retries2 - INTEGER 800 This value influences the timeout of an alive TCP connection, 801 when RTO retransmissions remain unacknowledged. 802 Given a value of N, a hypothetical TCP connection following 803 exponential backoff with an initial RTO of TCP_RTO_MIN would 804 retransmit N times before killing the connection at the (N+1)th RTO. 805 806 The default value of 15 yields a hypothetical timeout of 924.6 807 seconds and is a lower bound for the effective timeout. 808 TCP will effectively time out at the first RTO which exceeds the 809 hypothetical timeout. 810 If tcp_rto_max_ms is decreased, it is recommended to also 811 change tcp_retries2. 812 813 RFC 1122 recommends at least 100 seconds for the timeout, 814 which corresponds to a value of at least 8. 815 816tcp_rfc1337 - BOOLEAN 817 If enabled, the TCP stack behaves conforming to RFC1337. If unset, 818 we are not conforming to RFC, but prevent TCP TIME_WAIT 819 assassination. 820 821 Possible values: 822 823 - 0 (disabled) 824 - 1 (enabled) 825 826 Default: 0 (disabled) 827 828tcp_rmem - vector of 3 INTEGERs: min, default, max 829 min: Minimal size of receive buffer used by TCP sockets. 830 It is guaranteed to each TCP socket, even under moderate memory 831 pressure. 832 833 Default: 4K 834 835 default: initial size of receive buffer used by TCP sockets. 836 This value overrides net.core.rmem_default used by other protocols. 837 Default: 131072 bytes. 838 This value results in initial window of 65535. 839 840 max: maximal size of receive buffer allowed for automatically 841 selected receiver buffers for TCP socket. 842 Calling setsockopt() with SO_RCVBUF disables 843 automatic tuning of that socket's receive buffer size, in which 844 case this value is ignored. 845 Default: between 131072 and 32MB, depending on RAM size. 846 847tcp_sack - BOOLEAN 848 Enable select acknowledgments (SACKS). 849 850 Possible values: 851 852 - 0 (disabled) 853 - 1 (enabled) 854 855 Default: 1 (enabled) 856 857tcp_comp_sack_rtt_percent - INTEGER 858 Percentage of SRTT used for the compressed SACK feature. 859 See tcp_comp_sack_nr, tcp_comp_sack_delay_ns, tcp_comp_sack_slack_ns. 860 861 Possible values : 1 - 1000 862 863 Default : 33 % 864 865tcp_comp_sack_delay_ns - LONG INTEGER 866 TCP tries to reduce number of SACK sent, using a timer based 867 on tcp_comp_sack_rtt_percent of SRTT, capped by this sysctl 868 in nano seconds. 869 The default is 1ms, based on TSO autosizing period. 870 871 Default : 1,000,000 ns (1 ms) 872 873tcp_comp_sack_slack_ns - LONG INTEGER 874 This sysctl control the slack used when arming the 875 timer used by SACK compression. This gives extra time 876 for small RTT flows, and reduces system overhead by allowing 877 opportunistic reduction of timer interrupts. 878 879 Default : 100,000 ns (100 us) 880 881tcp_comp_sack_nr - INTEGER 882 Max number of SACK that can be compressed. 883 Using 0 disables SACK compression. 884 885 Default : 44 886 887tcp_backlog_ack_defer - BOOLEAN 888 If enabled, user thread processing socket backlog tries sending 889 one ACK for the whole queue. This helps to avoid potential 890 long latencies at end of a TCP socket syscall. 891 892 Possible values: 893 894 - 0 (disabled) 895 - 1 (enabled) 896 897 Default: 1 (enabled) 898 899tcp_slow_start_after_idle - BOOLEAN 900 If enabled, provide RFC2861 behavior and time out the congestion 901 window after an idle period. An idle period is defined at 902 the current RTO. If unset, the congestion window will not 903 be timed out after an idle period. 904 905 Possible values: 906 907 - 0 (disabled) 908 - 1 (enabled) 909 910 Default: 1 (enabled) 911 912tcp_stdurg - BOOLEAN 913 Use the Host requirements interpretation of the TCP urgent pointer field. 914 Most hosts use the older BSD interpretation, so if enabled, 915 Linux might not communicate correctly with them. 916 917 Possible values: 918 919 - 0 (disabled) 920 - 1 (enabled) 921 922 Default: 0 (disabled) 923 924tcp_synack_retries - INTEGER 925 Number of times SYNACKs for a passive TCP connection attempt will 926 be retransmitted. Should not be higher than 255. Default value 927 is 5, which corresponds to 31seconds till the last retransmission 928 with the current initial RTO of 1second. With this the final timeout 929 for a passive TCP connection will happen after 63seconds. 930 931tcp_syncookies - INTEGER 932 Only valid when the kernel was compiled with CONFIG_SYN_COOKIES 933 Send out syncookies when the syn backlog queue of a socket 934 overflows. This is to prevent against the common 'SYN flood attack' 935 Default: 1 936 937 Note, that syncookies is fallback facility. 938 It MUST NOT be used to help highly loaded servers to stand 939 against legal connection rate. If you see SYN flood warnings 940 in your logs, but investigation shows that they occur 941 because of overload with legal connections, you should tune 942 another parameters until this warning disappear. 943 See: tcp_max_syn_backlog, tcp_synack_retries, tcp_abort_on_overflow. 944 945 syncookies seriously violate TCP protocol, do not allow 946 to use TCP extensions, can result in serious degradation 947 of some services (f.e. SMTP relaying), visible not by you, 948 but your clients and relays, contacting you. While you see 949 SYN flood warnings in logs not being really flooded, your server 950 is seriously misconfigured. 951 952 If you want to test which effects syncookies have to your 953 network connections you can set this knob to 2 to enable 954 unconditionally generation of syncookies. 955 956tcp_migrate_req - BOOLEAN 957 The incoming connection is tied to a specific listening socket when 958 the initial SYN packet is received during the three-way handshake. 959 When a listener is closed, in-flight request sockets during the 960 handshake and established sockets in the accept queue are aborted. 961 962 If the listener has SO_REUSEPORT enabled, other listeners on the 963 same port should have been able to accept such connections. This 964 option makes it possible to migrate such child sockets to another 965 listener after close() or shutdown(). 966 967 The BPF_SK_REUSEPORT_SELECT_OR_MIGRATE type of eBPF program should 968 usually be used to define the policy to pick an alive listener. 969 Otherwise, the kernel will randomly pick an alive listener only if 970 this option is enabled. 971 972 Note that migration between listeners with different settings may 973 crash applications. Let's say migration happens from listener A to 974 B, and only B has TCP_SAVE_SYN enabled. B cannot read SYN data from 975 the requests migrated from A. To avoid such a situation, cancel 976 migration by returning SK_DROP in the type of eBPF program, or 977 disable this option. 978 979 Possible values: 980 981 - 0 (disabled) 982 - 1 (enabled) 983 984 Default: 0 (disabled) 985 986tcp_fastopen - INTEGER 987 Enable TCP Fast Open (RFC7413) to send and accept data in the opening 988 SYN packet. 989 990 The client support is enabled by flag 0x1 (on by default). The client 991 then must use sendmsg() or sendto() with the MSG_FASTOPEN flag, 992 rather than connect() to send data in SYN. 993 994 The server support is enabled by flag 0x2 (off by default). Then 995 either enable for all listeners with another flag (0x400) or 996 enable individual listeners via TCP_FASTOPEN socket option with 997 the option value being the length of the syn-data backlog. 998 999 The values (bitmap) are 1000 1001 ===== ======== ====================================================== 1002 0x1 (client) enables sending data in the opening SYN on the client. 1003 0x2 (server) enables the server support, i.e., allowing data in 1004 a SYN packet to be accepted and passed to the 1005 application before 3-way handshake finishes. 1006 0x4 (client) send data in the opening SYN regardless of cookie 1007 availability and without a cookie option. 1008 0x200 (server) accept data-in-SYN w/o any cookie option present. 1009 0x400 (server) enable all listeners to support Fast Open by 1010 default without explicit TCP_FASTOPEN socket option. 1011 ===== ======== ====================================================== 1012 1013 Default: 0x1 1014 1015 Note that additional client or server features are only 1016 effective if the basic support (0x1 and 0x2) are enabled respectively. 1017 1018tcp_fastopen_blackhole_timeout_sec - INTEGER 1019 Initial time period in second to disable Fastopen on active TCP sockets 1020 when a TFO firewall blackhole issue happens. 1021 This time period will grow exponentially when more blackhole issues 1022 get detected right after Fastopen is re-enabled and will reset to 1023 initial value when the blackhole issue goes away. 1024 0 to disable the blackhole detection. 1025 1026 By default, it is set to 0 (feature is disabled). 1027 1028tcp_fastopen_key - list of comma separated 32-digit hexadecimal INTEGERs 1029 The list consists of a primary key and an optional backup key. The 1030 primary key is used for both creating and validating cookies, while the 1031 optional backup key is only used for validating cookies. The purpose of 1032 the backup key is to maximize TFO validation when keys are rotated. 1033 1034 A randomly chosen primary key may be configured by the kernel if 1035 the tcp_fastopen sysctl is set to 0x400 (see above), or if the 1036 TCP_FASTOPEN setsockopt() optname is set and a key has not been 1037 previously configured via sysctl. If keys are configured via 1038 setsockopt() by using the TCP_FASTOPEN_KEY optname, then those 1039 per-socket keys will be used instead of any keys that are specified via 1040 sysctl. 1041 1042 A key is specified as 4 8-digit hexadecimal integers which are separated 1043 by a '-' as: xxxxxxxx-xxxxxxxx-xxxxxxxx-xxxxxxxx. Leading zeros may be 1044 omitted. A primary and a backup key may be specified by separating them 1045 by a comma. If only one key is specified, it becomes the primary key and 1046 any previously configured backup keys are removed. 1047 1048tcp_syn_retries - INTEGER 1049 Number of times initial SYNs for an active TCP connection attempt 1050 will be retransmitted. Should not be higher than 127. Default value 1051 is 6, which corresponds to 67seconds (with tcp_syn_linear_timeouts = 4) 1052 till the last retransmission with the current initial RTO of 1second. 1053 With this the final timeout for an active TCP connection attempt 1054 will happen after 131seconds. 1055 1056tcp_timestamps - INTEGER 1057 Enable timestamps as defined in RFC1323. 1058 1059 - 0: Disabled. 1060 - 1: Enable timestamps as defined in RFC1323 and use random offset for 1061 each connection rather than only using the current time. 1062 - 2: Like 1, but without random offsets. 1063 1064 Default: 1 1065 1066tcp_min_tso_segs - INTEGER 1067 Minimal number of segments per TSO frame. 1068 1069 Since linux-3.12, TCP does an automatic sizing of TSO frames, 1070 depending on flow rate, instead of filling 64Kbytes packets. 1071 For specific usages, it's possible to force TCP to build big 1072 TSO frames. Note that TCP stack might split too big TSO packets 1073 if available window is too small. 1074 1075 Default: 2 1076 1077tcp_tso_rtt_log - INTEGER 1078 Adjustment of TSO packet sizes based on min_rtt 1079 1080 Starting from linux-5.18, TCP autosizing can be tweaked 1081 for flows having small RTT. 1082 1083 Old autosizing was splitting the pacing budget to send 1024 TSO 1084 per second. 1085 1086 tso_packet_size = sk->sk_pacing_rate / 1024; 1087 1088 With the new mechanism, we increase this TSO sizing using: 1089 1090 distance = min_rtt_usec / (2^tcp_tso_rtt_log) 1091 tso_packet_size += gso_max_size >> distance; 1092 1093 This means that flows between very close hosts can use bigger 1094 TSO packets, reducing their cpu costs. 1095 1096 If you want to use the old autosizing, set this sysctl to 0. 1097 1098 Default: 9 (2^9 = 512 usec) 1099 1100tcp_pacing_ss_ratio - INTEGER 1101 sk->sk_pacing_rate is set by TCP stack using a ratio applied 1102 to current rate. (current_rate = cwnd * mss / srtt) 1103 If TCP is in slow start, tcp_pacing_ss_ratio is applied 1104 to let TCP probe for bigger speeds, assuming cwnd can be 1105 doubled every other RTT. 1106 1107 Default: 200 1108 1109tcp_pacing_ca_ratio - INTEGER 1110 sk->sk_pacing_rate is set by TCP stack using a ratio applied 1111 to current rate. (current_rate = cwnd * mss / srtt) 1112 If TCP is in congestion avoidance phase, tcp_pacing_ca_ratio 1113 is applied to conservatively probe for bigger throughput. 1114 1115 Default: 120 1116 1117tcp_syn_linear_timeouts - INTEGER 1118 The number of times for an active TCP connection to retransmit SYNs with 1119 a linear backoff timeout before defaulting to an exponential backoff 1120 timeout. This has no effect on SYNACK at the passive TCP side. 1121 1122 With an initial RTO of 1 and tcp_syn_linear_timeouts = 4 we would 1123 expect SYN RTOs to be: 1, 1, 1, 1, 1, 2, 4, ... (4 linear timeouts, 1124 and the first exponential backoff using 2^0 * initial_RTO). 1125 Default: 4 1126 1127tcp_tso_win_divisor - INTEGER 1128 This allows control over what percentage of the congestion window 1129 can be consumed by a single TSO frame. 1130 The setting of this parameter is a choice between burstiness and 1131 building larger TSO frames. 1132 1133 Default: 3 1134 1135tcp_tw_reuse - INTEGER 1136 Enable reuse of TIME-WAIT sockets for new connections when it is 1137 safe from protocol viewpoint. 1138 1139 - 0 - disable 1140 - 1 - global enable 1141 - 2 - enable for loopback traffic only 1142 1143 It should not be changed without advice/request of technical 1144 experts. 1145 1146 Default: 2 1147 1148tcp_tw_reuse_delay - UNSIGNED INTEGER 1149 The delay in milliseconds before a TIME-WAIT socket can be reused by a 1150 new connection, if TIME-WAIT socket reuse is enabled. The actual reuse 1151 threshold is within [N, N+1] range, where N is the requested delay in 1152 milliseconds, to ensure the delay interval is never shorter than the 1153 configured value. 1154 1155 This setting contains an assumption about the other TCP timestamp clock 1156 tick interval. It should not be set to a value lower than the peer's 1157 clock tick for PAWS (Protection Against Wrapped Sequence numbers) 1158 mechanism work correctly for the reused connection. 1159 1160 Default: 1000 (milliseconds) 1161 1162tcp_window_scaling - BOOLEAN 1163 Enable window scaling as defined in RFC1323. 1164 1165 Possible values: 1166 1167 - 0 (disabled) 1168 - 1 (enabled) 1169 1170 Default: 1 (enabled) 1171 1172tcp_shrink_window - BOOLEAN 1173 This changes how the TCP receive window is calculated. 1174 1175 RFC 7323, section 2.4, says there are instances when a retracted 1176 window can be offered, and that TCP implementations MUST ensure 1177 that they handle a shrinking window, as specified in RFC 1122. 1178 1179 Possible values: 1180 1181 - 0 (disabled) - The window is never shrunk. 1182 - 1 (enabled) - The window is shrunk when necessary to remain within 1183 the memory limit set by autotuning (sk_rcvbuf). 1184 This only occurs if a non-zero receive window 1185 scaling factor is also in effect. 1186 1187 Default: 0 (disabled) 1188 1189tcp_wmem - vector of 3 INTEGERs: min, default, max 1190 min: Amount of memory reserved for send buffers for TCP sockets. 1191 Each TCP socket has rights to use it due to fact of its birth. 1192 1193 Default: 4K 1194 1195 default: initial size of send buffer used by TCP sockets. This 1196 value overrides net.core.wmem_default used by other protocols. 1197 1198 It is usually lower than net.core.wmem_default. 1199 1200 Default: 16K 1201 1202 max: Maximal amount of memory allowed for automatically tuned 1203 send buffers for TCP sockets. This value does not override 1204 net.core.wmem_max. Calling setsockopt() with SO_SNDBUF disables 1205 automatic tuning of that socket's send buffer size, in which case 1206 this value is ignored. 1207 1208 Default: between 64K and 4MB, depending on RAM size. 1209 1210tcp_notsent_lowat - UNSIGNED INTEGER 1211 A TCP socket can control the amount of unsent bytes in its write queue, 1212 thanks to TCP_NOTSENT_LOWAT socket option. poll()/select()/epoll() 1213 reports POLLOUT events if the amount of unsent bytes is below a per 1214 socket value, and if the write queue is not full. sendmsg() will 1215 also not add new buffers if the limit is hit. 1216 1217 This global variable controls the amount of unsent data for 1218 sockets not using TCP_NOTSENT_LOWAT. For these sockets, a change 1219 to the global variable has immediate effect. 1220 1221 Default: UINT_MAX (0xFFFFFFFF) 1222 1223tcp_workaround_signed_windows - BOOLEAN 1224 If enabled, assume no receipt of a window scaling option means the 1225 remote TCP is broken and treats the window as a signed quantity. 1226 If disabled, assume the remote TCP is not broken even if we do 1227 not receive a window scaling option from them. 1228 1229 Possible values: 1230 1231 - 0 (disabled) 1232 - 1 (enabled) 1233 1234 Default: 0 (disabled) 1235 1236tcp_thin_linear_timeouts - BOOLEAN 1237 Enable dynamic triggering of linear timeouts for thin streams. 1238 If enabled, a check is performed upon retransmission by timeout to 1239 determine if the stream is thin (less than 4 packets in flight). 1240 As long as the stream is found to be thin, up to 6 linear 1241 timeouts may be performed before exponential backoff mode is 1242 initiated. This improves retransmission latency for 1243 non-aggressive thin streams, often found to be time-dependent. 1244 For more information on thin streams, see 1245 Documentation/networking/tcp-thin.rst 1246 1247 Possible values: 1248 1249 - 0 (disabled) 1250 - 1 (enabled) 1251 1252 Default: 0 (disabled) 1253 1254tcp_limit_output_bytes - INTEGER 1255 Controls TCP Small Queue limit per tcp socket. 1256 TCP bulk sender tends to increase packets in flight until it 1257 gets losses notifications. With SNDBUF autotuning, this can 1258 result in a large amount of packets queued on the local machine 1259 (e.g.: qdiscs, CPU backlog, or device) hurting latency of other 1260 flows, for typical pfifo_fast qdiscs. tcp_limit_output_bytes 1261 limits the number of bytes on qdisc or device to reduce artificial 1262 RTT/cwnd and reduce bufferbloat. 1263 1264 Default: 4194304 (4 MB) 1265 1266tcp_challenge_ack_limit - INTEGER 1267 Limits number of Challenge ACK sent per second, as recommended 1268 in RFC 5961 (Improving TCP's Robustness to Blind In-Window Attacks) 1269 Note that this per netns rate limit can allow some side channel 1270 attacks and probably should not be enabled. 1271 TCP stack implements per TCP socket limits anyway. 1272 Default: INT_MAX (unlimited) 1273 1274tcp_ehash_entries - INTEGER 1275 Show the number of hash buckets for TCP sockets in the current 1276 networking namespace. 1277 1278 A negative value means the networking namespace does not own its 1279 hash buckets and shares the initial networking namespace's one. 1280 1281tcp_child_ehash_entries - INTEGER 1282 Control the number of hash buckets for TCP sockets in the child 1283 networking namespace, which must be set before clone() or unshare(). 1284 1285 If the value is not 0, the kernel uses a value rounded up to 2^n 1286 as the actual hash bucket size. 0 is a special value, meaning 1287 the child networking namespace will share the initial networking 1288 namespace's hash buckets. 1289 1290 Note that the child will use the global one in case the kernel 1291 fails to allocate enough memory. In addition, the global hash 1292 buckets are spread over available NUMA nodes, but the allocation 1293 of the child hash table depends on the current process's NUMA 1294 policy, which could result in performance differences. 1295 1296 Note also that the default value of tcp_max_tw_buckets and 1297 tcp_max_syn_backlog depend on the hash bucket size. 1298 1299 Possible values: 0, 2^n (n: 0 - 24 (16Mi)) 1300 1301 Default: 0 1302 1303tcp_plb_enabled - BOOLEAN 1304 If enabled and the underlying congestion control (e.g. DCTCP) supports 1305 and enables PLB feature, TCP PLB (Protective Load Balancing) is 1306 enabled. PLB is described in the following paper: 1307 https://doi.org/10.1145/3544216.3544226. Based on PLB parameters, 1308 upon sensing sustained congestion, TCP triggers a change in 1309 flow label field for outgoing IPv6 packets. A change in flow label 1310 field potentially changes the path of outgoing packets for switches 1311 that use ECMP/WCMP for routing. 1312 1313 PLB changes socket txhash which results in a change in IPv6 Flow Label 1314 field, and currently no-op for IPv4 headers. It is possible 1315 to apply PLB for IPv4 with other network header fields (e.g. TCP 1316 or IPv4 options) or using encapsulation where outer header is used 1317 by switches to determine next hop. In either case, further host 1318 and switch side changes will be needed. 1319 1320 If enabled, PLB assumes that congestion signal (e.g. ECN) is made 1321 available and used by congestion control module to estimate a 1322 congestion measure (e.g. ce_ratio). PLB needs a congestion measure to 1323 make repathing decisions. 1324 1325 Possible values: 1326 1327 - 0 (disabled) 1328 - 1 (enabled) 1329 1330 Default: 0 (disabled) 1331 1332tcp_plb_idle_rehash_rounds - INTEGER 1333 Number of consecutive congested rounds (RTT) seen after which 1334 a rehash can be performed, given there are no packets in flight. 1335 This is referred to as M in PLB paper: 1336 https://doi.org/10.1145/3544216.3544226. 1337 1338 Possible Values: 0 - 31 1339 1340 Default: 3 1341 1342tcp_plb_rehash_rounds - INTEGER 1343 Number of consecutive congested rounds (RTT) seen after which 1344 a forced rehash can be performed. Be careful when setting this 1345 parameter, as a small value increases the risk of retransmissions. 1346 This is referred to as N in PLB paper: 1347 https://doi.org/10.1145/3544216.3544226. 1348 1349 Possible Values: 0 - 31 1350 1351 Default: 12 1352 1353tcp_plb_suspend_rto_sec - INTEGER 1354 Time, in seconds, to suspend PLB in event of an RTO. In order to avoid 1355 having PLB repath onto a connectivity "black hole", after an RTO a TCP 1356 connection suspends PLB repathing for a random duration between 1x and 1357 2x of this parameter. Randomness is added to avoid concurrent rehashing 1358 of multiple TCP connections. This should be set corresponding to the 1359 amount of time it takes to repair a failed link. 1360 1361 Possible Values: 0 - 255 1362 1363 Default: 60 1364 1365tcp_plb_cong_thresh - INTEGER 1366 Fraction of packets marked with congestion over a round (RTT) to 1367 tag that round as congested. This is referred to as K in the PLB paper: 1368 https://doi.org/10.1145/3544216.3544226. 1369 1370 The 0-1 fraction range is mapped to 0-256 range to avoid floating 1371 point operations. For example, 128 means that if at least 50% of 1372 the packets in a round were marked as congested then the round 1373 will be tagged as congested. 1374 1375 Setting threshold to 0 means that PLB repaths every RTT regardless 1376 of congestion. This is not intended behavior for PLB and should be 1377 used only for experimentation purpose. 1378 1379 Possible Values: 0 - 256 1380 1381 Default: 128 1382 1383tcp_pingpong_thresh - INTEGER 1384 The number of estimated data replies sent for estimated incoming data 1385 requests that must happen before TCP considers that a connection is a 1386 "ping-pong" (request-response) connection for which delayed 1387 acknowledgments can provide benefits. 1388 1389 This threshold is 1 by default, but some applications may need a higher 1390 threshold for optimal performance. 1391 1392 Possible Values: 1 - 255 1393 1394 Default: 1 1395 1396tcp_rto_min_us - INTEGER 1397 Minimal TCP retransmission timeout (in microseconds). Note that the 1398 rto_min route option has the highest precedence for configuring this 1399 setting, followed by the TCP_BPF_RTO_MIN and TCP_RTO_MIN_US socket 1400 options, followed by this tcp_rto_min_us sysctl. 1401 1402 The recommended practice is to use a value less or equal to 200000 1403 microseconds. 1404 1405 Possible Values: 1 - INT_MAX 1406 1407 Default: 200000 1408 1409tcp_rto_max_ms - INTEGER 1410 Maximal TCP retransmission timeout (in ms). 1411 Note that TCP_RTO_MAX_MS socket option has higher precedence. 1412 1413 When changing tcp_rto_max_ms, it is important to understand 1414 that tcp_retries2 might need a change. 1415 1416 Possible Values: 1000 - 120,000 1417 1418 Default: 120,000 1419 1420UDP variables 1421============= 1422 1423udp_l3mdev_accept - BOOLEAN 1424 Enabling this option allows a "global" bound socket to work 1425 across L3 master domains (e.g., VRFs) with packets capable of 1426 being received regardless of the L3 domain in which they 1427 originated. Only valid when the kernel was compiled with 1428 CONFIG_NET_L3_MASTER_DEV. 1429 1430 Possible values: 1431 1432 - 0 (disabled) 1433 - 1 (enabled) 1434 1435 Default: 0 (disabled) 1436 1437udp_mem - vector of 3 INTEGERs: min, pressure, max 1438 Number of pages allowed for queueing by all UDP sockets. 1439 1440 min: Number of pages allowed for queueing by all UDP sockets. 1441 1442 pressure: This value was introduced to follow format of tcp_mem. 1443 1444 max: This value was introduced to follow format of tcp_mem. 1445 1446 Default is calculated at boot time from amount of available memory. 1447 1448udp_rmem_min - INTEGER 1449 Minimal size of receive buffer used by UDP sockets in moderation. 1450 Each UDP socket is able to use the size for receiving data, even if 1451 total pages of UDP sockets exceed udp_mem pressure. The unit is byte. 1452 1453 Default: 4K 1454 1455udp_wmem_min - INTEGER 1456 UDP does not have tx memory accounting and this tunable has no effect. 1457 1458udp_hash_entries - INTEGER 1459 Show the number of hash buckets for UDP sockets in the current 1460 networking namespace. 1461 1462 A negative value means the networking namespace does not own its 1463 hash buckets and shares the initial networking namespace's one. 1464 1465udp_child_hash_entries - INTEGER 1466 Control the number of hash buckets for UDP sockets in the child 1467 networking namespace, which must be set before clone() or unshare(). 1468 1469 If the value is not 0, the kernel uses a value rounded up to 2^n 1470 as the actual hash bucket size. 0 is a special value, meaning 1471 the child networking namespace will share the initial networking 1472 namespace's hash buckets. 1473 1474 Note that the child will use the global one in case the kernel 1475 fails to allocate enough memory. In addition, the global hash 1476 buckets are spread over available NUMA nodes, but the allocation 1477 of the child hash table depends on the current process's NUMA 1478 policy, which could result in performance differences. 1479 1480 Possible values: 0, 2^n (n: 7 (128) - 16 (64K)) 1481 1482 Default: 0 1483 1484 1485RAW variables 1486============= 1487 1488raw_l3mdev_accept - BOOLEAN 1489 Enabling this option allows a "global" bound socket to work 1490 across L3 master domains (e.g., VRFs) with packets capable of 1491 being received regardless of the L3 domain in which they 1492 originated. Only valid when the kernel was compiled with 1493 CONFIG_NET_L3_MASTER_DEV. 1494 1495 Possible values: 1496 1497 - 0 (disabled) 1498 - 1 (enabled) 1499 1500 Default: 1 (enabled) 1501 1502CIPSOv4 Variables 1503================= 1504 1505cipso_cache_enable - BOOLEAN 1506 If enabled, enable additions to and lookups from the CIPSO label mapping 1507 cache. If disabled, additions are ignored and lookups always result in a 1508 miss. However, regardless of the setting the cache is still 1509 invalidated when required when means you can safely toggle this on and 1510 off and the cache will always be "safe". 1511 1512 Possible values: 1513 1514 - 0 (disabled) 1515 - 1 (enabled) 1516 1517 Default: 1 (enabled) 1518 1519cipso_cache_bucket_size - INTEGER 1520 The CIPSO label cache consists of a fixed size hash table with each 1521 hash bucket containing a number of cache entries. This variable limits 1522 the number of entries in each hash bucket; the larger the value is, the 1523 more CIPSO label mappings that can be cached. When the number of 1524 entries in a given hash bucket reaches this limit adding new entries 1525 causes the oldest entry in the bucket to be removed to make room. 1526 1527 Default: 10 1528 1529cipso_rbm_optfmt - BOOLEAN 1530 Enable the "Optimized Tag 1 Format" as defined in section 3.4.2.6 of 1531 the CIPSO draft specification (see Documentation/netlabel for details). 1532 This means that when set the CIPSO tag will be padded with empty 1533 categories in order to make the packet data 32-bit aligned. 1534 1535 Possible values: 1536 1537 - 0 (disabled) 1538 - 1 (enabled) 1539 1540 Default: 0 (disabled) 1541 1542cipso_rbm_strictvalid - BOOLEAN 1543 If enabled, do a very strict check of the CIPSO option when 1544 ip_options_compile() is called. If disabled, relax the checks done during 1545 ip_options_compile(). Either way is "safe" as errors are caught else 1546 where in the CIPSO processing code but setting this to 0 (False) should 1547 result in less work (i.e. it should be faster) but could cause problems 1548 with other implementations that require strict checking. 1549 1550 Possible values: 1551 1552 - 0 (disabled) 1553 - 1 (enabled) 1554 1555 Default: 0 (disabled) 1556 1557IP Variables 1558============ 1559 1560ip_local_port_range - 2 INTEGERS 1561 Defines the local port range that is used by TCP and UDP to 1562 choose the local port. The first number is the first, the 1563 second the last local port number. 1564 If possible, it is better these numbers have different parity 1565 (one even and one odd value). 1566 Must be greater than or equal to ip_unprivileged_port_start. 1567 The default values are 32768 and 60999 respectively. 1568 1569ip_local_reserved_ports - list of comma separated ranges 1570 Specify the ports which are reserved for known third-party 1571 applications. These ports will not be used by automatic port 1572 assignments (e.g. when calling connect() or bind() with port 1573 number 0). Explicit port allocation behavior is unchanged. 1574 1575 The format used for both input and output is a comma separated 1576 list of ranges (e.g. "1,2-4,10-10" for ports 1, 2, 3, 4 and 1577 10). Writing to the file will clear all previously reserved 1578 ports and update the current list with the one given in the 1579 input. 1580 1581 Note that ip_local_port_range and ip_local_reserved_ports 1582 settings are independent and both are considered by the kernel 1583 when determining which ports are available for automatic port 1584 assignments. 1585 1586 You can reserve ports which are not in the current 1587 ip_local_port_range, e.g.:: 1588 1589 $ cat /proc/sys/net/ipv4/ip_local_port_range 1590 32000 60999 1591 $ cat /proc/sys/net/ipv4/ip_local_reserved_ports 1592 8080,9148 1593 1594 although this is redundant. However such a setting is useful 1595 if later the port range is changed to a value that will 1596 include the reserved ports. Also keep in mind, that overlapping 1597 of these ranges may affect probability of selecting ephemeral 1598 ports which are right after block of reserved ports. 1599 1600 Default: Empty 1601 1602ip_unprivileged_port_start - INTEGER 1603 This is a per-namespace sysctl. It defines the first 1604 unprivileged port in the network namespace. Privileged ports 1605 require root or CAP_NET_BIND_SERVICE in order to bind to them. 1606 To disable all privileged ports, set this to 0. They must not 1607 overlap with the ip_local_port_range. 1608 1609 Default: 1024 1610 1611ip_nonlocal_bind - BOOLEAN 1612 If enabled, allows processes to bind() to non-local IP addresses, 1613 which can be quite useful - but may break some applications. 1614 1615 Possible values: 1616 1617 - 0 (disabled) 1618 - 1 (enabled) 1619 1620 Default: 0 (disabled) 1621 1622ip_autobind_reuse - BOOLEAN 1623 By default, bind() does not select the ports automatically even if 1624 the new socket and all sockets bound to the port have SO_REUSEADDR. 1625 ip_autobind_reuse allows bind() to reuse the port and this is useful 1626 when you use bind()+connect(), but may break some applications. 1627 The preferred solution is to use IP_BIND_ADDRESS_NO_PORT and this 1628 option should only be set by experts. 1629 1630 Possible values: 1631 1632 - 0 (disabled) 1633 - 1 (enabled) 1634 1635 Default: 0 (disabled) 1636 1637ip_dynaddr - INTEGER 1638 If set non-zero, enables support for dynamic addresses. 1639 If set to a non-zero value larger than 1, a kernel log 1640 message will be printed when dynamic address rewriting 1641 occurs. 1642 1643 Default: 0 1644 1645ip_early_demux - BOOLEAN 1646 Optimize input packet processing down to one demux for 1647 certain kinds of local sockets. Currently we only do this 1648 for established TCP and connected UDP sockets. 1649 1650 It may add an additional cost for pure routing workloads that 1651 reduces overall throughput, in such case you should disable it. 1652 1653 Possible values: 1654 1655 - 0 (disabled) 1656 - 1 (enabled) 1657 1658 Default: 1 (enabled) 1659 1660ping_group_range - 2 INTEGERS 1661 Restrict ICMP_PROTO datagram sockets to users in the group range. 1662 The default is "1 0", meaning, that nobody (not even root) may 1663 create ping sockets. Setting it to "100 100" would grant permissions 1664 to the single group. "0 4294967294" would enable it for the world, "100 1665 4294967294" would enable it for the users, but not daemons. 1666 1667tcp_early_demux - BOOLEAN 1668 Enable early demux for established TCP sockets. 1669 1670 Possible values: 1671 1672 - 0 (disabled) 1673 - 1 (enabled) 1674 1675 Default: 1 (enabled) 1676 1677udp_early_demux - BOOLEAN 1678 Enable early demux for connected UDP sockets. Disable this if 1679 your system could experience more unconnected load. 1680 1681 Possible values: 1682 1683 - 0 (disabled) 1684 - 1 (enabled) 1685 1686 Default: 1 (enabled) 1687 1688icmp_echo_ignore_all - BOOLEAN 1689 If enabled, then the kernel will ignore all ICMP ECHO 1690 requests sent to it. 1691 1692 Possible values: 1693 1694 - 0 (disabled) 1695 - 1 (enabled) 1696 1697 Default: 0 (disabled) 1698 1699icmp_echo_enable_probe - BOOLEAN 1700 If enabled, then the kernel will respond to RFC 8335 PROBE 1701 requests sent to it. 1702 1703 Possible values: 1704 1705 - 0 (disabled) 1706 - 1 (enabled) 1707 1708 Default: 0 (disabled) 1709 1710icmp_echo_ignore_broadcasts - BOOLEAN 1711 If enabled, then the kernel will ignore all ICMP ECHO and 1712 TIMESTAMP requests sent to it via broadcast/multicast. 1713 1714 Possible values: 1715 1716 - 0 (disabled) 1717 - 1 (enabled) 1718 1719 Default: 1 (enabled) 1720 1721icmp_ratelimit - INTEGER 1722 Limit the maximal rates for sending ICMP packets whose type matches 1723 icmp_ratemask (see below) to specific targets. 1724 0 to disable any limiting, 1725 otherwise the minimal space between responses in milliseconds. 1726 Note that another sysctl, icmp_msgs_per_sec limits the number 1727 of ICMP packets sent on all targets. 1728 1729 Default: 1000 1730 1731icmp_msgs_per_sec - INTEGER 1732 Limit maximal number of ICMP packets sent per second from this host. 1733 Only messages whose type matches icmp_ratemask (see below) are 1734 controlled by this limit. For security reasons, the precise count 1735 of messages per second is randomized. 1736 1737 Default: 1000 1738 1739icmp_msgs_burst - INTEGER 1740 icmp_msgs_per_sec controls number of ICMP packets sent per second, 1741 while icmp_msgs_burst controls the burst size of these packets. 1742 For security reasons, the precise burst size is randomized. 1743 1744 Default: 50 1745 1746icmp_ratemask - INTEGER 1747 Mask made of ICMP types for which rates are being limited. 1748 1749 Significant bits: IHGFEDCBA9876543210 1750 1751 Default mask: 0000001100000011000 (6168) 1752 1753 Bit definitions (see include/linux/icmp.h): 1754 1755 = ========================= 1756 0 Echo Reply 1757 3 Destination Unreachable [1]_ 1758 4 Source Quench [1]_ 1759 5 Redirect 1760 8 Echo Request 1761 B Time Exceeded [1]_ 1762 C Parameter Problem [1]_ 1763 D Timestamp Request 1764 E Timestamp Reply 1765 F Info Request 1766 G Info Reply 1767 H Address Mask Request 1768 I Address Mask Reply 1769 = ========================= 1770 1771 .. [1] These are rate limited by default (see default mask above) 1772 1773icmp_ignore_bogus_error_responses - BOOLEAN 1774 Some routers violate RFC1122 by sending bogus responses to broadcast 1775 frames. Such violations are normally logged via a kernel warning. 1776 If enabled, the kernel will not give such warnings, which 1777 will avoid log file clutter. 1778 1779 Possible values: 1780 1781 - 0 (disabled) 1782 - 1 (enabled) 1783 1784 Default: 1 (enabled) 1785 1786icmp_errors_use_inbound_ifaddr - BOOLEAN 1787 1788 If disabled, icmp error messages are sent with the primary address of 1789 the exiting interface. 1790 1791 If enabled, the message will be sent with the primary address of 1792 the interface that received the packet that caused the icmp error. 1793 This is the behaviour many network administrators will expect from 1794 a router. And it can make debugging complicated network layouts 1795 much easier. 1796 1797 Note that if no primary address exists for the interface selected, 1798 then the primary address of the first non-loopback interface that 1799 has one will be used regardless of this setting. 1800 1801 Possible values: 1802 1803 - 0 (disabled) 1804 - 1 (enabled) 1805 1806 Default: 0 (disabled) 1807 1808icmp_errors_extension_mask - UNSIGNED INTEGER 1809 Bitmask of ICMP extensions to append to ICMPv4 error messages 1810 ("Destination Unreachable", "Time Exceeded" and "Parameter Problem"). 1811 The original datagram is trimmed / padded to 128 bytes in order to be 1812 compatible with applications that do not comply with RFC 4884. 1813 1814 Possible extensions are: 1815 1816 ==== ============================================================== 1817 0x01 Incoming IP interface information according to RFC 5837. 1818 Extension will include the index, IPv4 address (if present), 1819 name and MTU of the IP interface that received the datagram 1820 which elicited the ICMP error. 1821 ==== ============================================================== 1822 1823 Default: 0x00 (no extensions) 1824 1825igmp_max_memberships - INTEGER 1826 Change the maximum number of multicast groups we can subscribe to. 1827 Default: 20 1828 1829 Theoretical maximum value is bounded by having to send a membership 1830 report in a single datagram (i.e. the report can't span multiple 1831 datagrams, or risk confusing the switch and leaving groups you don't 1832 intend to). 1833 1834 The number of supported groups 'M' is bounded by the number of group 1835 report entries you can fit into a single datagram of 65535 bytes. 1836 1837 M = 65536-sizeof (ip header)/(sizeof(Group record)) 1838 1839 Group records are variable length, with a minimum of 12 bytes. 1840 So net.ipv4.igmp_max_memberships should not be set higher than: 1841 1842 (65536-24) / 12 = 5459 1843 1844 The value 5459 assumes no IP header options, so in practice 1845 this number may be lower. 1846 1847igmp_max_msf - INTEGER 1848 Maximum number of addresses allowed in the source filter list for a 1849 multicast group. 1850 1851 Default: 10 1852 1853igmp_qrv - INTEGER 1854 Controls the IGMP query robustness variable (see RFC2236 8.1). 1855 1856 Default: 2 (as specified by RFC2236 8.1) 1857 1858 Minimum: 1 (as specified by RFC6636 4.5) 1859 1860force_igmp_version - INTEGER 1861 - 0 - (default) No enforcement of a IGMP version, IGMPv1/v2 fallback 1862 allowed. Will back to IGMPv3 mode again if all IGMPv1/v2 Querier 1863 Present timer expires. 1864 - 1 - Enforce to use IGMP version 1. Will also reply IGMPv1 report if 1865 receive IGMPv2/v3 query. 1866 - 2 - Enforce to use IGMP version 2. Will fallback to IGMPv1 if receive 1867 IGMPv1 query message. Will reply report if receive IGMPv3 query. 1868 - 3 - Enforce to use IGMP version 3. The same react with default 0. 1869 1870 .. note:: 1871 1872 this is not the same with force_mld_version because IGMPv3 RFC3376 1873 Security Considerations does not have clear description that we could 1874 ignore other version messages completely as MLDv2 RFC3810. So make 1875 this value as default 0 is recommended. 1876 1877``conf/interface/*`` 1878 changes special settings per interface (where 1879 interface" is the name of your network interface) 1880 1881``conf/all/*`` 1882 is special, changes the settings for all interfaces 1883 1884log_martians - BOOLEAN 1885 Log packets with impossible addresses to kernel log. 1886 log_martians for the interface will be enabled if at least one of 1887 conf/{all,interface}/log_martians is set to TRUE, 1888 it will be disabled otherwise 1889 1890accept_redirects - BOOLEAN 1891 Accept ICMP redirect messages. 1892 accept_redirects for the interface will be enabled if: 1893 1894 - both conf/{all,interface}/accept_redirects are TRUE in the case 1895 forwarding for the interface is enabled 1896 1897 or 1898 1899 - at least one of conf/{all,interface}/accept_redirects is TRUE in the 1900 case forwarding for the interface is disabled 1901 1902 accept_redirects for the interface will be disabled otherwise 1903 1904 default: 1905 1906 - TRUE (host) 1907 - FALSE (router) 1908 1909forwarding - BOOLEAN 1910 Enable IP forwarding on this interface. This controls whether packets 1911 received _on_ this interface can be forwarded. 1912 1913mc_forwarding - BOOLEAN 1914 Do multicast routing. The kernel needs to be compiled with CONFIG_MROUTE 1915 and a multicast routing daemon is required. 1916 conf/all/mc_forwarding must also be set to TRUE to enable multicast 1917 routing for the interface 1918 1919medium_id - INTEGER 1920 Integer value used to differentiate the devices by the medium they 1921 are attached to. Two devices can have different id values when 1922 the broadcast packets are received only on one of them. 1923 The default value 0 means that the device is the only interface 1924 to its medium, value of -1 means that medium is not known. 1925 1926 Currently, it is used to change the proxy_arp behavior: 1927 the proxy_arp feature is enabled for packets forwarded between 1928 two devices attached to different media. 1929 1930proxy_arp - BOOLEAN 1931 Do proxy arp. 1932 1933 proxy_arp for the interface will be enabled if at least one of 1934 conf/{all,interface}/proxy_arp is set to TRUE, 1935 it will be disabled otherwise 1936 1937proxy_arp_pvlan - BOOLEAN 1938 Private VLAN proxy arp. 1939 1940 Basically allow proxy arp replies back to the same interface 1941 (from which the ARP request/solicitation was received). 1942 1943 This is done to support (ethernet) switch features, like RFC 1944 3069, where the individual ports are NOT allowed to 1945 communicate with each other, but they are allowed to talk to 1946 the upstream router. As described in RFC 3069, it is possible 1947 to allow these hosts to communicate through the upstream 1948 router by proxy_arp'ing. Don't need to be used together with 1949 proxy_arp. 1950 1951 This technology is known by different names: 1952 1953 - In RFC 3069 it is called VLAN Aggregation. 1954 - Cisco and Allied Telesyn call it Private VLAN. 1955 - Hewlett-Packard call it Source-Port filtering or port-isolation. 1956 - Ericsson call it MAC-Forced Forwarding (RFC Draft). 1957 1958proxy_delay - INTEGER 1959 Delay proxy response. 1960 1961 Delay response to a neighbor solicitation when proxy_arp 1962 or proxy_ndp is enabled. A random value between [0, proxy_delay) 1963 will be chosen, setting to zero means reply with no delay. 1964 Value in jiffies. Defaults to 80. 1965 1966shared_media - BOOLEAN 1967 Send(router) or accept(host) RFC1620 shared media redirects. 1968 Overrides secure_redirects. 1969 1970 shared_media for the interface will be enabled if at least one of 1971 conf/{all,interface}/shared_media is set to TRUE, 1972 it will be disabled otherwise 1973 1974 default TRUE 1975 1976secure_redirects - BOOLEAN 1977 Accept ICMP redirect messages only to gateways listed in the 1978 interface's current gateway list. Even if disabled, RFC1122 redirect 1979 rules still apply. 1980 1981 Overridden by shared_media. 1982 1983 secure_redirects for the interface will be enabled if at least one of 1984 conf/{all,interface}/secure_redirects is set to TRUE, 1985 it will be disabled otherwise 1986 1987 default TRUE 1988 1989send_redirects - BOOLEAN 1990 Send redirects, if router. 1991 1992 send_redirects for the interface will be enabled if at least one of 1993 conf/{all,interface}/send_redirects is set to TRUE, 1994 it will be disabled otherwise 1995 1996 Default: TRUE 1997 1998bootp_relay - BOOLEAN 1999 Accept packets with source address 0.b.c.d destined 2000 not to this host as local ones. It is supposed, that 2001 BOOTP relay daemon will catch and forward such packets. 2002 conf/all/bootp_relay must also be set to TRUE to enable BOOTP relay 2003 for the interface 2004 2005 default FALSE 2006 2007 Not Implemented Yet. 2008 2009accept_source_route - BOOLEAN 2010 Accept packets with SRR option. 2011 conf/all/accept_source_route must also be set to TRUE to accept packets 2012 with SRR option on the interface 2013 2014 default 2015 2016 - TRUE (router) 2017 - FALSE (host) 2018 2019accept_local - BOOLEAN 2020 Accept packets with local source addresses. In combination with 2021 suitable routing, this can be used to direct packets between two 2022 local interfaces over the wire and have them accepted properly. 2023 default FALSE 2024 2025route_localnet - BOOLEAN 2026 Do not consider loopback addresses as martian source or destination 2027 while routing. This enables the use of 127/8 for local routing purposes. 2028 2029 default FALSE 2030 2031rp_filter - INTEGER 2032 - 0 - No source validation. 2033 - 1 - Strict mode as defined in RFC3704 Strict Reverse Path 2034 Each incoming packet is tested against the FIB and if the interface 2035 is not the best reverse path the packet check will fail. 2036 By default failed packets are discarded. 2037 - 2 - Loose mode as defined in RFC3704 Loose Reverse Path 2038 Each incoming packet's source address is also tested against the FIB 2039 and if the source address is not reachable via any interface 2040 the packet check will fail. 2041 2042 Current recommended practice in RFC3704 is to enable strict mode 2043 to prevent IP spoofing from DDos attacks. If using asymmetric routing 2044 or other complicated routing, then loose mode is recommended. 2045 2046 The max value from conf/{all,interface}/rp_filter is used 2047 when doing source validation on the {interface}. 2048 2049 Default value is 0. Note that some distributions enable it 2050 in startup scripts. 2051 2052src_valid_mark - BOOLEAN 2053 - 0 - The fwmark of the packet is not included in reverse path 2054 route lookup. This allows for asymmetric routing configurations 2055 utilizing the fwmark in only one direction, e.g., transparent 2056 proxying. 2057 2058 - 1 - The fwmark of the packet is included in reverse path route 2059 lookup. This permits rp_filter to function when the fwmark is 2060 used for routing traffic in both directions. 2061 2062 This setting also affects the utilization of fmwark when 2063 performing source address selection for ICMP replies, or 2064 determining addresses stored for the IPOPT_TS_TSANDADDR and 2065 IPOPT_RR IP options. 2066 2067 The max value from conf/{all,interface}/src_valid_mark is used. 2068 2069 Default value is 0. 2070 2071arp_filter - BOOLEAN 2072 - 1 - Allows you to have multiple network interfaces on the same 2073 subnet, and have the ARPs for each interface be answered 2074 based on whether or not the kernel would route a packet from 2075 the ARP'd IP out that interface (therefore you must use source 2076 based routing for this to work). In other words it allows control 2077 of which cards (usually 1) will respond to an arp request. 2078 2079 - 0 - (default) The kernel can respond to arp requests with addresses 2080 from other interfaces. This may seem wrong but it usually makes 2081 sense, because it increases the chance of successful communication. 2082 IP addresses are owned by the complete host on Linux, not by 2083 particular interfaces. Only for more complex setups like load- 2084 balancing, does this behaviour cause problems. 2085 2086 arp_filter for the interface will be enabled if at least one of 2087 conf/{all,interface}/arp_filter is set to TRUE, 2088 it will be disabled otherwise 2089 2090arp_announce - INTEGER 2091 Define different restriction levels for announcing the local 2092 source IP address from IP packets in ARP requests sent on 2093 interface: 2094 2095 - 0 - (default) Use any local address, configured on any interface 2096 - 1 - Try to avoid local addresses that are not in the target's 2097 subnet for this interface. This mode is useful when target 2098 hosts reachable via this interface require the source IP 2099 address in ARP requests to be part of their logical network 2100 configured on the receiving interface. When we generate the 2101 request we will check all our subnets that include the 2102 target IP and will preserve the source address if it is from 2103 such subnet. If there is no such subnet we select source 2104 address according to the rules for level 2. 2105 - 2 - Always use the best local address for this target. 2106 In this mode we ignore the source address in the IP packet 2107 and try to select local address that we prefer for talks with 2108 the target host. Such local address is selected by looking 2109 for primary IP addresses on all our subnets on the outgoing 2110 interface that include the target IP address. If no suitable 2111 local address is found we select the first local address 2112 we have on the outgoing interface or on all other interfaces, 2113 with the hope we will receive reply for our request and 2114 even sometimes no matter the source IP address we announce. 2115 2116 The max value from conf/{all,interface}/arp_announce is used. 2117 2118 Increasing the restriction level gives more chance for 2119 receiving answer from the resolved target while decreasing 2120 the level announces more valid sender's information. 2121 2122arp_ignore - INTEGER 2123 Define different modes for sending replies in response to 2124 received ARP requests that resolve local target IP addresses: 2125 2126 - 0 - (default): reply for any local target IP address, configured 2127 on any interface 2128 - 1 - reply only if the target IP address is local address 2129 configured on the incoming interface 2130 - 2 - reply only if the target IP address is local address 2131 configured on the incoming interface and both with the 2132 sender's IP address are part from same subnet on this interface 2133 - 3 - do not reply for local addresses configured with scope host, 2134 only resolutions for global and link addresses are replied 2135 - 4-7 - reserved 2136 - 8 - do not reply for all local addresses 2137 2138 The max value from conf/{all,interface}/arp_ignore is used 2139 when ARP request is received on the {interface} 2140 2141arp_notify - BOOLEAN 2142 Define mode for notification of address and device changes. 2143 2144 == ========================================================== 2145 0 (default): do nothing 2146 1 Generate gratuitous arp requests when device is brought up 2147 or hardware address changes. 2148 == ========================================================== 2149 2150arp_accept - INTEGER 2151 Define behavior for accepting gratuitous ARP (garp) frames from devices 2152 that are not already present in the ARP table: 2153 2154 - 0 - don't create new entries in the ARP table 2155 - 1 - create new entries in the ARP table 2156 - 2 - create new entries only if the source IP address is in the same 2157 subnet as an address configured on the interface that received the 2158 garp message. 2159 2160 Both replies and requests type gratuitous arp will trigger the 2161 ARP table to be updated, if this setting is on. 2162 2163 If the ARP table already contains the IP address of the 2164 gratuitous arp frame, the arp table will be updated regardless 2165 if this setting is on or off. 2166 2167arp_evict_nocarrier - BOOLEAN 2168 Clears the ARP cache on NOCARRIER events. This option is important for 2169 wireless devices where the ARP cache should not be cleared when roaming 2170 between access points on the same network. In most cases this should 2171 remain as the default (1). 2172 2173 Possible values: 2174 2175 - 0 (disabled) - Do not clear ARP cache on NOCARRIER events 2176 - 1 (enabled) - Clear the ARP cache on NOCARRIER events 2177 2178 Default: 1 (enabled) 2179 2180mcast_solicit - INTEGER 2181 The maximum number of multicast probes in INCOMPLETE state, 2182 when the associated hardware address is unknown. Defaults 2183 to 3. 2184 2185ucast_solicit - INTEGER 2186 The maximum number of unicast probes in PROBE state, when 2187 the hardware address is being reconfirmed. Defaults to 3. 2188 2189app_solicit - INTEGER 2190 The maximum number of probes to send to the user space ARP daemon 2191 via netlink before dropping back to multicast probes (see 2192 mcast_resolicit). Defaults to 0. 2193 2194mcast_resolicit - INTEGER 2195 The maximum number of multicast probes after unicast and 2196 app probes in PROBE state. Defaults to 0. 2197 2198disable_policy - BOOLEAN 2199 Disable IPSEC policy (SPD) for this interface 2200 2201 Possible values: 2202 2203 - 0 (disabled) 2204 - 1 (enabled) 2205 2206 Default: 0 (disabled) 2207 2208disable_xfrm - BOOLEAN 2209 Disable IPSEC encryption on this interface, whatever the policy 2210 2211 Possible values: 2212 2213 - 0 (disabled) 2214 - 1 (enabled) 2215 2216 Default: 0 (disabled) 2217 2218igmpv2_unsolicited_report_interval - INTEGER 2219 The interval in milliseconds in which the next unsolicited 2220 IGMPv1 or IGMPv2 report retransmit will take place. 2221 2222 Default: 10000 (10 seconds) 2223 2224igmpv3_unsolicited_report_interval - INTEGER 2225 The interval in milliseconds in which the next unsolicited 2226 IGMPv3 report retransmit will take place. 2227 2228 Default: 1000 (1 seconds) 2229 2230ignore_routes_with_linkdown - BOOLEAN 2231 Ignore routes whose link is down when performing a FIB lookup. 2232 2233 Possible values: 2234 2235 - 0 (disabled) 2236 - 1 (enabled) 2237 2238 Default: 0 (disabled) 2239 2240promote_secondaries - BOOLEAN 2241 When a primary IP address is removed from this interface 2242 promote a corresponding secondary IP address instead of 2243 removing all the corresponding secondary IP addresses. 2244 2245 Possible values: 2246 2247 - 0 (disabled) 2248 - 1 (enabled) 2249 2250 Default: 0 (disabled) 2251 2252drop_unicast_in_l2_multicast - BOOLEAN 2253 Drop any unicast IP packets that are received in link-layer 2254 multicast (or broadcast) frames. 2255 2256 This behavior (for multicast) is actually a SHOULD in RFC 2257 1122, but is disabled by default for compatibility reasons. 2258 2259 Possible values: 2260 2261 - 0 (disabled) 2262 - 1 (enabled) 2263 2264 Default: 0 (disabled) 2265 2266drop_gratuitous_arp - BOOLEAN 2267 Drop all gratuitous ARP frames, for example if there's a known 2268 good ARP proxy on the network and such frames need not be used 2269 (or in the case of 802.11, must not be used to prevent attacks.) 2270 2271 Possible values: 2272 2273 - 0 (disabled) 2274 - 1 (enabled) 2275 2276 Default: 0 (disabled) 2277 2278 2279tag - INTEGER 2280 Allows you to write a number, which can be used as required. 2281 2282 Default value is 0. 2283 2284xfrm4_gc_thresh - INTEGER 2285 (Obsolete since linux-4.14) 2286 The threshold at which we will start garbage collecting for IPv4 2287 destination cache entries. At twice this value the system will 2288 refuse new allocations. 2289 2290igmp_link_local_mcast_reports - BOOLEAN 2291 Enable IGMP reports for link local multicast groups in the 2292 224.0.0.X range. 2293 2294 Default TRUE 2295 2296Alexey Kuznetsov. 2297kuznet@ms2.inr.ac.ru 2298 2299Updated by: 2300 2301- Andi Kleen 2302 ak@muc.de 2303- Nicolas Delon 2304 delon.nicolas@wanadoo.fr 2305 2306 2307 2308 2309/proc/sys/net/ipv6/* Variables 2310============================== 2311 2312IPv6 has no global variables such as tcp_*. tcp_* settings under ipv4/ also 2313apply to IPv6 [XXX?]. 2314 2315bindv6only - BOOLEAN 2316 Default value for IPV6_V6ONLY socket option, 2317 which restricts use of the IPv6 socket to IPv6 communication 2318 only. 2319 2320 Possible values: 2321 2322 - 0 (disabled) - enable IPv4-mapped address feature 2323 - 1 (enabled) - disable IPv4-mapped address feature 2324 2325 Default: 0 (disabled) 2326 2327flowlabel_consistency - BOOLEAN 2328 Protect the consistency (and unicity) of flow label. 2329 You have to disable it to use IPV6_FL_F_REFLECT flag on the 2330 flow label manager. 2331 2332 Possible values: 2333 2334 - 0 (disabled) 2335 - 1 (enabled) 2336 2337 Default: 1 (enabled) 2338 2339auto_flowlabels - INTEGER 2340 Automatically generate flow labels based on a flow hash of the 2341 packet. This allows intermediate devices, such as routers, to 2342 identify packet flows for mechanisms like Equal Cost Multipath 2343 Routing (see RFC 6438). 2344 2345 = =========================================================== 2346 0 automatic flow labels are completely disabled 2347 1 automatic flow labels are enabled by default, they can be 2348 disabled on a per socket basis using the IPV6_AUTOFLOWLABEL 2349 socket option 2350 2 automatic flow labels are allowed, they may be enabled on a 2351 per socket basis using the IPV6_AUTOFLOWLABEL socket option 2352 3 automatic flow labels are enabled and enforced, they cannot 2353 be disabled by the socket option 2354 = =========================================================== 2355 2356 Default: 1 2357 2358flowlabel_state_ranges - BOOLEAN 2359 Split the flow label number space into two ranges. 0-0x7FFFF is 2360 reserved for the IPv6 flow manager facility, 0x80000-0xFFFFF 2361 is reserved for stateless flow labels as described in RFC6437. 2362 2363 Possible values: 2364 2365 - 0 (disabled) 2366 - 1 (enabled) 2367 2368 Default: 1 (enabled) 2369 2370 2371flowlabel_reflect - INTEGER 2372 Control flow label reflection. Needed for Path MTU 2373 Discovery to work with Equal Cost Multipath Routing in anycast 2374 environments. See RFC 7690 and: 2375 https://tools.ietf.org/html/draft-wang-6man-flow-label-reflection-01 2376 2377 This is a bitmask. 2378 2379 - 1: enabled for established flows 2380 2381 Note that this prevents automatic flowlabel changes, as done 2382 in "tcp: change IPv6 flow-label upon receiving spurious retransmission" 2383 and "tcp: Change txhash on every SYN and RTO retransmit" 2384 2385 - 2: enabled for TCP RESET packets (no active listener) 2386 If set, a RST packet sent in response to a SYN packet on a closed 2387 port will reflect the incoming flow label. 2388 2389 - 4: enabled for ICMPv6 echo reply messages. 2390 2391 Default: 0 2392 2393fib_multipath_hash_policy - INTEGER 2394 Controls which hash policy to use for multipath routes. 2395 2396 Default: 0 (Layer 3) 2397 2398 Possible values: 2399 2400 - 0 - Layer 3 (source and destination addresses plus flow label) 2401 - 1 - Layer 4 (standard 5-tuple) 2402 - 2 - Layer 3 or inner Layer 3 if present 2403 - 3 - Custom multipath hash. Fields used for multipath hash calculation 2404 are determined by fib_multipath_hash_fields sysctl 2405 2406fib_multipath_hash_fields - UNSIGNED INTEGER 2407 When fib_multipath_hash_policy is set to 3 (custom multipath hash), the 2408 fields used for multipath hash calculation are determined by this 2409 sysctl. 2410 2411 This value is a bitmask which enables various fields for multipath hash 2412 calculation. 2413 2414 Possible fields are: 2415 2416 ====== ============================ 2417 0x0001 Source IP address 2418 0x0002 Destination IP address 2419 0x0004 IP protocol 2420 0x0008 Flow Label 2421 0x0010 Source port 2422 0x0020 Destination port 2423 0x0040 Inner source IP address 2424 0x0080 Inner destination IP address 2425 0x0100 Inner IP protocol 2426 0x0200 Inner Flow Label 2427 0x0400 Inner source port 2428 0x0800 Inner destination port 2429 ====== ============================ 2430 2431 Default: 0x0007 (source IP, destination IP and IP protocol) 2432 2433anycast_src_echo_reply - BOOLEAN 2434 Controls the use of anycast addresses as source addresses for ICMPv6 2435 echo reply 2436 2437 Possible values: 2438 2439 - 0 (disabled) 2440 - 1 (enabled) 2441 2442 Default: 0 (disabled) 2443 2444 2445idgen_delay - INTEGER 2446 Controls the delay in seconds after which time to retry 2447 privacy stable address generation if a DAD conflict is 2448 detected. 2449 2450 Default: 1 (as specified in RFC7217) 2451 2452idgen_retries - INTEGER 2453 Controls the number of retries to generate a stable privacy 2454 address if a DAD conflict is detected. 2455 2456 Default: 3 (as specified in RFC7217) 2457 2458mld_qrv - INTEGER 2459 Controls the MLD query robustness variable (see RFC3810 9.1). 2460 2461 Default: 2 (as specified by RFC3810 9.1) 2462 2463 Minimum: 1 (as specified by RFC6636 4.5) 2464 2465max_dst_opts_number - INTEGER 2466 Maximum number of non-padding TLVs allowed in a Destination 2467 options extension header. If this value is less than zero 2468 then unknown options are disallowed and the number of known 2469 TLVs allowed is the absolute value of this number. 2470 2471 Default: 8 2472 2473max_hbh_opts_number - INTEGER 2474 Maximum number of non-padding TLVs allowed in a Hop-by-Hop 2475 options extension header. If this value is less than zero 2476 then unknown options are disallowed and the number of known 2477 TLVs allowed is the absolute value of this number. 2478 2479 Default: 8 2480 2481max_dst_opts_length - INTEGER 2482 Maximum length allowed for a Destination options extension 2483 header. 2484 2485 Default: INT_MAX (unlimited) 2486 2487max_hbh_length - INTEGER 2488 Maximum length allowed for a Hop-by-Hop options extension 2489 header. 2490 2491 Default: INT_MAX (unlimited) 2492 2493skip_notify_on_dev_down - BOOLEAN 2494 Controls whether an RTM_DELROUTE message is generated for routes 2495 removed when a device is taken down or deleted. IPv4 does not 2496 generate this message; IPv6 does by default. Setting this sysctl 2497 to true skips the message, making IPv4 and IPv6 on par in relying 2498 on userspace caches to track link events and evict routes. 2499 2500 Possible values: 2501 2502 - 0 (disabled) - generate the message 2503 - 1 (enabled) - skip generating the message 2504 2505 Default: 0 (disabled) 2506 2507nexthop_compat_mode - BOOLEAN 2508 New nexthop API provides a means for managing nexthops independent of 2509 prefixes. Backwards compatibility with old route format is enabled by 2510 default which means route dumps and notifications contain the new 2511 nexthop attribute but also the full, expanded nexthop definition. 2512 Further, updates or deletes of a nexthop configuration generate route 2513 notifications for each fib entry using the nexthop. Once a system 2514 understands the new API, this sysctl can be disabled to achieve full 2515 performance benefits of the new API by disabling the nexthop expansion 2516 and extraneous notifications. 2517 2518 Note that as a backward-compatible mode, dumping of modern features 2519 might be incomplete or wrong. For example, resilient groups will not be 2520 shown as such, but rather as just a list of next hops. Also weights that 2521 do not fit into 8 bits will show incorrectly. 2522 2523 Default: true (backward compat mode) 2524 2525fib_notify_on_flag_change - INTEGER 2526 Whether to emit RTM_NEWROUTE notifications whenever RTM_F_OFFLOAD/ 2527 RTM_F_TRAP/RTM_F_OFFLOAD_FAILED flags are changed. 2528 2529 After installing a route to the kernel, user space receives an 2530 acknowledgment, which means the route was installed in the kernel, 2531 but not necessarily in hardware. 2532 It is also possible for a route already installed in hardware to change 2533 its action and therefore its flags. For example, a host route that is 2534 trapping packets can be "promoted" to perform decapsulation following 2535 the installation of an IPinIP/VXLAN tunnel. 2536 The notifications will indicate to user-space the state of the route. 2537 2538 Default: 0 (Do not emit notifications.) 2539 2540 Possible values: 2541 2542 - 0 - Do not emit notifications. 2543 - 1 - Emit notifications. 2544 - 2 - Emit notifications only for RTM_F_OFFLOAD_FAILED flag change. 2545 2546ioam6_id - INTEGER 2547 Define the IOAM id of this node. Uses only 24 bits out of 32 in total. 2548 2549 Possible value range: 2550 2551 - Min: 0 2552 - Max: 0xFFFFFF 2553 2554 Default: 0xFFFFFF 2555 2556ioam6_id_wide - LONG INTEGER 2557 Define the wide IOAM id of this node. Uses only 56 bits out of 64 in 2558 total. Can be different from ioam6_id. 2559 2560 Possible value range: 2561 2562 - Min: 0 2563 - Max: 0xFFFFFFFFFFFFFF 2564 2565 Default: 0xFFFFFFFFFFFFFF 2566 2567IPv6 Fragmentation: 2568 2569ip6frag_high_thresh - INTEGER 2570 Maximum memory used to reassemble IPv6 fragments. When 2571 ip6frag_high_thresh bytes of memory is allocated for this purpose, 2572 the fragment handler will toss packets until ip6frag_low_thresh 2573 is reached. 2574 2575ip6frag_low_thresh - INTEGER 2576 See ip6frag_high_thresh 2577 2578ip6frag_time - INTEGER 2579 Time in seconds to keep an IPv6 fragment in memory. 2580 2581``conf/default/*``: 2582 Change the interface-specific default settings. 2583 2584 These settings would be used during creating new interfaces. 2585 2586 2587``conf/all/*``: 2588 Change all the interface-specific settings. 2589 2590 [XXX: Other special features than forwarding?] 2591 2592conf/all/disable_ipv6 - BOOLEAN 2593 Changing this value is same as changing ``conf/default/disable_ipv6`` 2594 setting and also all per-interface ``disable_ipv6`` settings to the same 2595 value. 2596 2597 Reading this value does not have any particular meaning. It does not say 2598 whether IPv6 support is enabled or disabled. Returned value can be 1 2599 also in the case when some interface has ``disable_ipv6`` set to 0 and 2600 has configured IPv6 addresses. 2601 2602conf/all/forwarding - BOOLEAN 2603 Enable global IPv6 forwarding between all interfaces. 2604 2605 IPv4 and IPv6 work differently here; the ``force_forwarding`` flag must 2606 be used to control which interfaces may forward packets. 2607 2608 This also sets all interfaces' Host/Router setting 2609 'forwarding' to the specified value. See below for details. 2610 2611 This referred to as global forwarding. 2612 2613proxy_ndp - BOOLEAN 2614 Do proxy ndp. 2615 2616 Possible values: 2617 2618 - 0 (disabled) 2619 - 1 (enabled) 2620 2621 Default: 0 (disabled) 2622 2623force_forwarding - BOOLEAN 2624 Enable forwarding on this interface only -- regardless of the setting on 2625 ``conf/all/forwarding``. When setting ``conf.all.forwarding`` to 0, 2626 the ``force_forwarding`` flag will be reset on all interfaces. 2627 2628fwmark_reflect - BOOLEAN 2629 Controls the fwmark of kernel-generated IPv6 reply packets that are not 2630 associated with a socket for example, TCP RSTs or ICMPv6 echo replies). 2631 If disabled, these packets have a fwmark of zero. If enabled, they have the 2632 fwmark of the packet they are replying to. 2633 2634 Possible values: 2635 2636 - 0 (disabled) 2637 - 1 (enabled) 2638 2639 Default: 0 (disabled) 2640 2641``conf/interface/*``: 2642 Change special settings per interface. 2643 2644 The functional behaviour for certain settings is different 2645 depending on whether local forwarding is enabled or not. 2646 2647accept_ra - INTEGER 2648 Accept Router Advertisements; autoconfigure using them. 2649 2650 It also determines whether or not to transmit Router 2651 Solicitations. If and only if the functional setting is to 2652 accept Router Advertisements, Router Solicitations will be 2653 transmitted. 2654 2655 Possible values are: 2656 2657 == =========================================================== 2658 0 Do not accept Router Advertisements. 2659 1 Accept Router Advertisements if forwarding is disabled. 2660 2 Overrule forwarding behaviour. Accept Router Advertisements 2661 even if forwarding is enabled. 2662 == =========================================================== 2663 2664 Functional default: 2665 2666 - enabled if local forwarding is disabled. 2667 - disabled if local forwarding is enabled. 2668 2669accept_ra_defrtr - BOOLEAN 2670 Learn default router in Router Advertisement. 2671 2672 Functional default: 2673 2674 - enabled if accept_ra is enabled. 2675 - disabled if accept_ra is disabled. 2676 2677ra_defrtr_metric - UNSIGNED INTEGER 2678 Route metric for default route learned in Router Advertisement. This value 2679 will be assigned as metric for the default route learned via IPv6 Router 2680 Advertisement. Takes affect only if accept_ra_defrtr is enabled. 2681 2682 Possible values: 2683 1 to 0xFFFFFFFF 2684 2685 Default: IP6_RT_PRIO_USER i.e. 1024. 2686 2687accept_ra_from_local - BOOLEAN 2688 Accept RA with source-address that is found on local machine 2689 if the RA is otherwise proper and able to be accepted. 2690 2691 Default is to NOT accept these as it may be an un-intended 2692 network loop. 2693 2694 Functional default: 2695 2696 - enabled if accept_ra_from_local is enabled 2697 on a specific interface. 2698 - disabled if accept_ra_from_local is disabled 2699 on a specific interface. 2700 2701accept_ra_min_hop_limit - INTEGER 2702 Minimum hop limit Information in Router Advertisement. 2703 2704 Hop limit Information in Router Advertisement less than this 2705 variable shall be ignored. 2706 2707 Default: 1 2708 2709accept_ra_min_lft - INTEGER 2710 Minimum acceptable lifetime value in Router Advertisement. 2711 2712 RA sections with a lifetime less than this value shall be 2713 ignored. Zero lifetimes stay unaffected. 2714 2715 Default: 0 2716 2717accept_ra_pinfo - BOOLEAN 2718 Learn Prefix Information in Router Advertisement. 2719 2720 Functional default: 2721 2722 - enabled if accept_ra is enabled. 2723 - disabled if accept_ra is disabled. 2724 2725ra_honor_pio_life - BOOLEAN 2726 Whether to use RFC4862 Section 5.5.3e to determine the valid 2727 lifetime of an address matching a prefix sent in a Router 2728 Advertisement Prefix Information Option. 2729 2730 Possible values: 2731 2732 - 0 (disabled) - RFC4862 section 5.5.3e is used to determine 2733 the valid lifetime of the address. 2734 - 1 (enabled) - the PIO valid lifetime will always be honored. 2735 2736 Default: 0 (disabled) 2737 2738ra_honor_pio_pflag - BOOLEAN 2739 The Prefix Information Option P-flag indicates the network can 2740 allocate a unique IPv6 prefix per client using DHCPv6-PD. 2741 This sysctl can be enabled when a userspace DHCPv6-PD client 2742 is running to cause the P-flag to take effect: i.e. the 2743 P-flag suppresses any effects of the A-flag within the same 2744 PIO. For a given PIO, P=1 and A=1 is treated as A=0. 2745 2746 Possible values: 2747 2748 - 0 (disabled) - the P-flag is ignored. 2749 - 1 (enabled) - the P-flag will disable SLAAC autoconfiguration 2750 for the given Prefix Information Option. 2751 2752 Default: 0 (disabled) 2753 2754accept_ra_rt_info_min_plen - INTEGER 2755 Minimum prefix length of Route Information in RA. 2756 2757 Route Information w/ prefix smaller than this variable shall 2758 be ignored. 2759 2760 Functional default: 2761 2762 * 0 if accept_ra_rtr_pref is enabled. 2763 * -1 if accept_ra_rtr_pref is disabled. 2764 2765accept_ra_rt_info_max_plen - INTEGER 2766 Maximum prefix length of Route Information in RA. 2767 2768 Route Information w/ prefix larger than this variable shall 2769 be ignored. 2770 2771 Functional default: 2772 2773 * 0 if accept_ra_rtr_pref is enabled. 2774 * -1 if accept_ra_rtr_pref is disabled. 2775 2776accept_ra_rtr_pref - BOOLEAN 2777 Accept Router Preference in RA. 2778 2779 Functional default: 2780 2781 - enabled if accept_ra is enabled. 2782 - disabled if accept_ra is disabled. 2783 2784accept_ra_mtu - BOOLEAN 2785 Apply the MTU value specified in RA option 5 (RFC4861). If 2786 disabled, the MTU specified in the RA will be ignored. 2787 2788 Functional default: 2789 2790 - enabled if accept_ra is enabled. 2791 - disabled if accept_ra is disabled. 2792 2793accept_redirects - BOOLEAN 2794 Accept Redirects. 2795 2796 Functional default: 2797 2798 - enabled if local forwarding is disabled. 2799 - disabled if local forwarding is enabled. 2800 2801accept_source_route - INTEGER 2802 Accept source routing (routing extension header). 2803 2804 - >= 0: Accept only routing header type 2. 2805 - < 0: Do not accept routing header. 2806 2807 Default: 0 2808 2809autoconf - BOOLEAN 2810 Autoconfigure addresses using Prefix Information in Router 2811 Advertisements. 2812 2813 Functional default: 2814 2815 - enabled if accept_ra_pinfo is enabled. 2816 - disabled if accept_ra_pinfo is disabled. 2817 2818dad_transmits - INTEGER 2819 The amount of Duplicate Address Detection probes to send. 2820 2821 Default: 1 2822 2823forwarding - INTEGER 2824 Configure interface-specific Host/Router behaviour. 2825 2826 .. note:: 2827 2828 It is recommended to have the same setting on all 2829 interfaces; mixed router/host scenarios are rather uncommon. 2830 2831 Possible values are: 2832 2833 - 0 Forwarding disabled 2834 - 1 Forwarding enabled 2835 2836 **FALSE (0)**: 2837 2838 By default, Host behaviour is assumed. This means: 2839 2840 1. IsRouter flag is not set in Neighbour Advertisements. 2841 2. If accept_ra is TRUE (default), transmit Router 2842 Solicitations. 2843 3. If accept_ra is TRUE (default), accept Router 2844 Advertisements (and do autoconfiguration). 2845 4. If accept_redirects is TRUE (default), accept Redirects. 2846 2847 **TRUE (1)**: 2848 2849 If local forwarding is enabled, Router behaviour is assumed. 2850 This means exactly the reverse from the above: 2851 2852 1. IsRouter flag is set in Neighbour Advertisements. 2853 2. Router Solicitations are not sent unless accept_ra is 2. 2854 3. Router Advertisements are ignored unless accept_ra is 2. 2855 4. Redirects are ignored. 2856 2857 Default: 0 (disabled) if global forwarding is disabled (default), 2858 otherwise 1 (enabled). 2859 2860hop_limit - INTEGER 2861 Default Hop Limit to set. 2862 2863 Default: 64 2864 2865mtu - INTEGER 2866 Default Maximum Transfer Unit 2867 2868 Default: 1280 (IPv6 required minimum) 2869 2870ip_nonlocal_bind - BOOLEAN 2871 If enabled, allows processes to bind() to non-local IPv6 addresses, 2872 which can be quite useful - but may break some applications. 2873 2874 Possible values: 2875 2876 - 0 (disabled) 2877 - 1 (enabled) 2878 2879 Default: 0 (disabled) 2880 2881router_probe_interval - INTEGER 2882 Minimum interval (in seconds) between Router Probing described 2883 in RFC4191. 2884 2885 Default: 60 2886 2887router_solicitation_delay - INTEGER 2888 Number of seconds to wait after interface is brought up 2889 before sending Router Solicitations. 2890 2891 Default: 1 2892 2893router_solicitation_interval - INTEGER 2894 Number of seconds to wait between Router Solicitations. 2895 2896 Default: 4 2897 2898router_solicitations - INTEGER 2899 Number of Router Solicitations to send until assuming no 2900 routers are present. 2901 2902 Default: 3 2903 2904use_oif_addrs_only - BOOLEAN 2905 When enabled, the candidate source addresses for destinations 2906 routed via this interface are restricted to the set of addresses 2907 configured on this interface (vis. RFC 6724, section 4). 2908 2909 Possible values: 2910 2911 - 0 (disabled) 2912 - 1 (enabled) 2913 2914 Default: 0 (disabled) 2915 2916use_tempaddr - INTEGER 2917 Preference for Privacy Extensions (RFC3041). 2918 2919 * <= 0 : disable Privacy Extensions 2920 * == 1 : enable Privacy Extensions, but prefer public 2921 addresses over temporary addresses. 2922 * > 1 : enable Privacy Extensions and prefer temporary 2923 addresses over public addresses. 2924 2925 Default: 2926 2927 * 0 (for most devices) 2928 * -1 (for point-to-point devices and loopback devices) 2929 2930temp_valid_lft - INTEGER 2931 valid lifetime (in seconds) for temporary addresses. If less than the 2932 minimum required lifetime (typically 5-7 seconds), temporary addresses 2933 will not be created. 2934 2935 Default: 172800 (2 days) 2936 2937temp_prefered_lft - INTEGER 2938 Preferred lifetime (in seconds) for temporary addresses. If 2939 temp_prefered_lft is less than the minimum required lifetime (typically 2940 5-7 seconds), the preferred lifetime is the minimum required. If 2941 temp_prefered_lft is greater than temp_valid_lft, the preferred lifetime 2942 is temp_valid_lft. 2943 2944 Default: 86400 (1 day) 2945 2946keep_addr_on_down - INTEGER 2947 Keep all IPv6 addresses on an interface down event. If set static 2948 global addresses with no expiration time are not flushed. 2949 2950 * >0 : enabled 2951 * 0 : system default 2952 * <0 : disabled 2953 2954 Default: 0 (addresses are removed) 2955 2956max_desync_factor - INTEGER 2957 Maximum value for DESYNC_FACTOR, which is a random value 2958 that ensures that clients don't synchronize with each 2959 other and generate new addresses at exactly the same time. 2960 value is in seconds. 2961 2962 Default: 600 2963 2964regen_min_advance - INTEGER 2965 How far in advance (in seconds), at minimum, to create a new temporary 2966 address before the current one is deprecated. This value is added to 2967 the amount of time that may be required for duplicate address detection 2968 to determine when to create a new address. Linux permits setting this 2969 value to less than the default of 2 seconds, but a value less than 2 2970 does not conform to RFC 8981. 2971 2972 Default: 2 2973 2974regen_max_retry - INTEGER 2975 Number of attempts before give up attempting to generate 2976 valid temporary addresses. 2977 2978 Default: 5 2979 2980max_addresses - INTEGER 2981 Maximum number of autoconfigured addresses per interface. Setting 2982 to zero disables the limitation. It is not recommended to set this 2983 value too large (or to zero) because it would be an easy way to 2984 crash the kernel by allowing too many addresses to be created. 2985 2986 Default: 16 2987 2988disable_ipv6 - BOOLEAN 2989 Disable IPv6 operation. If accept_dad is set to 2, this value 2990 will be dynamically set to TRUE if DAD fails for the link-local 2991 address. 2992 2993 Default: FALSE (enable IPv6 operation) 2994 2995 When this value is changed from 1 to 0 (IPv6 is being enabled), 2996 it will dynamically create a link-local address on the given 2997 interface and start Duplicate Address Detection, if necessary. 2998 2999 When this value is changed from 0 to 1 (IPv6 is being disabled), 3000 it will dynamically delete all addresses and routes on the given 3001 interface. From now on it will not possible to add addresses/routes 3002 to the selected interface. 3003 3004accept_dad - INTEGER 3005 Whether to accept DAD (Duplicate Address Detection). 3006 3007 == ============================================================== 3008 0 Disable DAD 3009 1 Enable DAD (default) 3010 2 Enable DAD, and disable IPv6 operation if MAC-based duplicate 3011 link-local address has been found. 3012 == ============================================================== 3013 3014 DAD operation and mode on a given interface will be selected according 3015 to the maximum value of conf/{all,interface}/accept_dad. 3016 3017force_tllao - BOOLEAN 3018 Enable sending the target link-layer address option even when 3019 responding to a unicast neighbor solicitation. 3020 3021 Default: FALSE 3022 3023 Quoting from RFC 2461, section 4.4, Target link-layer address: 3024 3025 "The option MUST be included for multicast solicitations in order to 3026 avoid infinite Neighbor Solicitation "recursion" when the peer node 3027 does not have a cache entry to return a Neighbor Advertisements 3028 message. When responding to unicast solicitations, the option can be 3029 omitted since the sender of the solicitation has the correct link- 3030 layer address; otherwise it would not have be able to send the unicast 3031 solicitation in the first place. However, including the link-layer 3032 address in this case adds little overhead and eliminates a potential 3033 race condition where the sender deletes the cached link-layer address 3034 prior to receiving a response to a previous solicitation." 3035 3036ndisc_notify - BOOLEAN 3037 Define mode for notification of address and device changes. 3038 3039 Possible values: 3040 3041 - 0 (disabled) - do nothing 3042 - 1 (enabled) - Generate unsolicited neighbour advertisements when device is brought 3043 up or hardware address changes. 3044 3045 Default: 0 (disabled) 3046 3047ndisc_tclass - INTEGER 3048 The IPv6 Traffic Class to use by default when sending IPv6 Neighbor 3049 Discovery (Router Solicitation, Router Advertisement, Neighbor 3050 Solicitation, Neighbor Advertisement, Redirect) messages. 3051 These 8 bits can be interpreted as 6 high order bits holding the DSCP 3052 value and 2 low order bits representing ECN (which you probably want 3053 to leave cleared). 3054 3055 * 0 - (default) 3056 3057ndisc_evict_nocarrier - BOOLEAN 3058 Clears the neighbor discovery table on NOCARRIER events. This option is 3059 important for wireless devices where the neighbor discovery cache should 3060 not be cleared when roaming between access points on the same network. 3061 In most cases this should remain as the default (1). 3062 3063 Possible values: 3064 3065 - 0 (disabled) - Do not clear neighbor discovery cache on NOCARRIER events. 3066 - 1 (enabled) - Clear neighbor discover cache on NOCARRIER events. 3067 3068 Default: 1 (enabled) 3069 3070mldv1_unsolicited_report_interval - INTEGER 3071 The interval in milliseconds in which the next unsolicited 3072 MLDv1 report retransmit will take place. 3073 3074 Default: 10000 (10 seconds) 3075 3076mldv2_unsolicited_report_interval - INTEGER 3077 The interval in milliseconds in which the next unsolicited 3078 MLDv2 report retransmit will take place. 3079 3080 Default: 1000 (1 second) 3081 3082force_mld_version - INTEGER 3083 * 0 - (default) No enforcement of a MLD version, MLDv1 fallback allowed 3084 * 1 - Enforce to use MLD version 1 3085 * 2 - Enforce to use MLD version 2 3086 3087suppress_frag_ndisc - INTEGER 3088 Control RFC 6980 (Security Implications of IPv6 Fragmentation 3089 with IPv6 Neighbor Discovery) behavior: 3090 3091 * 1 - (default) discard fragmented neighbor discovery packets 3092 * 0 - allow fragmented neighbor discovery packets 3093 3094optimistic_dad - BOOLEAN 3095 Whether to perform Optimistic Duplicate Address Detection (RFC 4429). 3096 3097 Optimistic Duplicate Address Detection for the interface will be enabled 3098 if at least one of conf/{all,interface}/optimistic_dad is set to 1, 3099 it will be disabled otherwise. 3100 3101 Possible values: 3102 3103 - 0 (disabled) 3104 - 1 (enabled) 3105 3106 Default: 0 (disabled) 3107 3108 3109use_optimistic - BOOLEAN 3110 If enabled, do not classify optimistic addresses as deprecated during 3111 source address selection. Preferred addresses will still be chosen 3112 before optimistic addresses, subject to other ranking in the source 3113 address selection algorithm. 3114 3115 This will be enabled if at least one of 3116 conf/{all,interface}/use_optimistic is set to 1, disabled otherwise. 3117 3118 Possible values: 3119 3120 - 0 (disabled) 3121 - 1 (enabled) 3122 3123 Default: 0 (disabled) 3124 3125stable_secret - IPv6 address 3126 This IPv6 address will be used as a secret to generate IPv6 3127 addresses for link-local addresses and autoconfigured 3128 ones. All addresses generated after setting this secret will 3129 be stable privacy ones by default. This can be changed via the 3130 addrgenmode ip-link. conf/default/stable_secret is used as the 3131 secret for the namespace, the interface specific ones can 3132 overwrite that. Writes to conf/all/stable_secret are refused. 3133 3134 It is recommended to generate this secret during installation 3135 of a system and keep it stable after that. 3136 3137 By default the stable secret is unset. 3138 3139addr_gen_mode - INTEGER 3140 Defines how link-local and autoconf addresses are generated. 3141 3142 = ================================================================= 3143 0 generate address based on EUI64 (default) 3144 1 do no generate a link-local address, use EUI64 for addresses 3145 generated from autoconf 3146 2 generate stable privacy addresses, using the secret from 3147 stable_secret (RFC7217) 3148 3 generate stable privacy addresses, using a random secret if unset 3149 = ================================================================= 3150 3151drop_unicast_in_l2_multicast - BOOLEAN 3152 Drop any unicast IPv6 packets that are received in link-layer 3153 multicast (or broadcast) frames. 3154 3155 Possible values: 3156 3157 - 0 (disabled) 3158 - 1 (enabled) 3159 3160 Default: 0 (disabled) 3161 3162drop_unsolicited_na - BOOLEAN 3163 Drop all unsolicited neighbor advertisements, for example if there's 3164 a known good NA proxy on the network and such frames need not be used 3165 (or in the case of 802.11, must not be used to prevent attacks.) 3166 3167 Possible values: 3168 3169 - 0 (disabled) 3170 - 1 (enabled) 3171 3172 Default: 0 (disabled). 3173 3174accept_untracked_na - INTEGER 3175 Define behavior for accepting neighbor advertisements from devices that 3176 are absent in the neighbor cache: 3177 3178 - 0 - (default) Do not accept unsolicited and untracked neighbor 3179 advertisements. 3180 3181 - 1 - Add a new neighbor cache entry in STALE state for routers on 3182 receiving a neighbor advertisement (either solicited or unsolicited) 3183 with target link-layer address option specified if no neighbor entry 3184 is already present for the advertised IPv6 address. Without this knob, 3185 NAs received for untracked addresses (absent in neighbor cache) are 3186 silently ignored. 3187 3188 This is as per router-side behavior documented in RFC9131. 3189 3190 This has lower precedence than drop_unsolicited_na. 3191 3192 This will optimize the return path for the initial off-link 3193 communication that is initiated by a directly connected host, by 3194 ensuring that the first-hop router which turns on this setting doesn't 3195 have to buffer the initial return packets to do neighbor-solicitation. 3196 The prerequisite is that the host is configured to send unsolicited 3197 neighbor advertisements on interface bringup. This setting should be 3198 used in conjunction with the ndisc_notify setting on the host to 3199 satisfy this prerequisite. 3200 3201 - 2 - Extend option (1) to add a new neighbor cache entry only if the 3202 source IP address is in the same subnet as an address configured on 3203 the interface that received the neighbor advertisement. 3204 3205enhanced_dad - BOOLEAN 3206 Include a nonce option in the IPv6 neighbor solicitation messages used for 3207 duplicate address detection per RFC7527. A received DAD NS will only signal 3208 a duplicate address if the nonce is different. This avoids any false 3209 detection of duplicates due to loopback of the NS messages that we send. 3210 The nonce option will be sent on an interface unless both of 3211 conf/{all,interface}/enhanced_dad are set to FALSE. 3212 3213 Possible values: 3214 3215 - 0 (disabled) 3216 - 1 (enabled) 3217 3218 Default: 1 (enabled) 3219 3220``icmp/*``: 3221=========== 3222 3223ratelimit - INTEGER 3224 Limit the maximal rates for sending ICMPv6 messages. 3225 3226 0 to disable any limiting, 3227 otherwise the minimal space between responses in milliseconds. 3228 3229 Default: 1000 3230 3231ratemask - list of comma separated ranges 3232 For ICMPv6 message types matching the ranges in the ratemask, limit 3233 the sending of the message according to ratelimit parameter. 3234 3235 The format used for both input and output is a comma separated 3236 list of ranges (e.g. "0-127,129" for ICMPv6 message type 0 to 127 and 3237 129). Writing to the file will clear all previous ranges of ICMPv6 3238 message types and update the current list with the input. 3239 3240 Refer to: https://www.iana.org/assignments/icmpv6-parameters/icmpv6-parameters.xhtml 3241 for numerical values of ICMPv6 message types, e.g. echo request is 128 3242 and echo reply is 129. 3243 3244 Default: 0-1,3-127 (rate limit ICMPv6 errors except Packet Too Big) 3245 3246echo_ignore_all - BOOLEAN 3247 If enabled, then the kernel will ignore all ICMP ECHO 3248 requests sent to it over the IPv6 protocol. 3249 3250 Possible values: 3251 3252 - 0 (disabled) 3253 - 1 (enabled) 3254 3255 Default: 0 (disabled) 3256 3257echo_ignore_multicast - BOOLEAN 3258 If enabled, then the kernel will ignore all ICMP ECHO 3259 requests sent to it over the IPv6 protocol via multicast. 3260 3261 Possible values: 3262 3263 - 0 (disabled) 3264 - 1 (enabled) 3265 3266 Default: 0 (disabled) 3267 3268echo_ignore_anycast - BOOLEAN 3269 If enabled, then the kernel will ignore all ICMP ECHO 3270 requests sent to it over the IPv6 protocol destined to anycast address. 3271 3272 Possible values: 3273 3274 - 0 (disabled) 3275 - 1 (enabled) 3276 3277 Default: 0 (disabled) 3278 3279error_anycast_as_unicast - BOOLEAN 3280 If enabled, then the kernel will respond with ICMP Errors 3281 resulting from requests sent to it over the IPv6 protocol destined 3282 to anycast address essentially treating anycast as unicast. 3283 3284 Possible values: 3285 3286 - 0 (disabled) 3287 - 1 (enabled) 3288 3289 Default: 0 (disabled) 3290 3291errors_extension_mask - UNSIGNED INTEGER 3292 Bitmask of ICMP extensions to append to ICMPv6 error messages 3293 ("Destination Unreachable" and "Time Exceeded"). The original datagram 3294 is trimmed / padded to 128 bytes in order to be compatible with 3295 applications that do not comply with RFC 4884. 3296 3297 Possible extensions are: 3298 3299 ==== ============================================================== 3300 0x01 Incoming IP interface information according to RFC 5837. 3301 Extension will include the index, IPv6 address (if present), 3302 name and MTU of the IP interface that received the datagram 3303 which elicited the ICMP error. 3304 ==== ============================================================== 3305 3306 Default: 0x00 (no extensions) 3307 3308xfrm6_gc_thresh - INTEGER 3309 (Obsolete since linux-4.14) 3310 The threshold at which we will start garbage collecting for IPv6 3311 destination cache entries. At twice this value the system will 3312 refuse new allocations. 3313 3314 3315IPv6 Update by: 3316Pekka Savola <pekkas@netcore.fi> 3317YOSHIFUJI Hideaki / USAGI Project <yoshfuji@linux-ipv6.org> 3318 3319 3320/proc/sys/net/bridge/* Variables: 3321================================= 3322 3323bridge-nf-call-arptables - BOOLEAN 3324 3325 Possible values: 3326 3327 - 0 (disabled) - disable this. 3328 - 1 (enabled) - pass bridged ARP traffic to arptables' FORWARD chain. 3329 3330 Default: 1 (enabled) 3331 3332bridge-nf-call-iptables - BOOLEAN 3333 3334 Possible values: 3335 3336 - 0 (disabled) - disable this. 3337 - 1 (enabled) - pass bridged IPv4 traffic to iptables' chains. 3338 3339 Default: 1 (enabled) 3340 3341bridge-nf-call-ip6tables - BOOLEAN 3342 3343 Possible values: 3344 3345 - 0 (disabled) - disable this. 3346 - 1 (enabled) - pass bridged IPv6 traffic to ip6tables' chains. 3347 3348 Default: 1 (enabled) 3349 3350bridge-nf-filter-vlan-tagged - BOOLEAN 3351 3352 Possible values: 3353 3354 - 0 (disabled) - disable this. 3355 - 1 (enabled) - pass bridged vlan-tagged ARP/IP/IPv6 traffic to {arp,ip,ip6}tables 3356 3357 Default: 0 (disabled) 3358 3359bridge-nf-filter-pppoe-tagged - BOOLEAN 3360 3361 Possible values: 3362 3363 - 0 (disabled) - disable this. 3364 - 1 (enabled) - pass bridged pppoe-tagged IP/IPv6 traffic to {ip,ip6}tables. 3365 3366 Default: 0 (disabled) 3367 3368bridge-nf-pass-vlan-input-dev - BOOLEAN 3369 - 1: if bridge-nf-filter-vlan-tagged is enabled, try to find a vlan 3370 interface on the bridge and set the netfilter input device to the 3371 vlan. This allows use of e.g. "iptables -i br0.1" and makes the 3372 REDIRECT target work with vlan-on-top-of-bridge interfaces. When no 3373 matching vlan interface is found, or this switch is off, the input 3374 device is set to the bridge interface. 3375 3376 - 0: disable bridge netfilter vlan interface lookup. 3377 3378 Default: 0 3379 3380``proc/sys/net/sctp/*`` Variables: 3381================================== 3382 3383addip_enable - BOOLEAN 3384 Enable or disable extension of Dynamic Address Reconfiguration 3385 (ADD-IP) functionality specified in RFC5061. This extension provides 3386 the ability to dynamically add and remove new addresses for the SCTP 3387 associations. 3388 3389 Possible values: 3390 3391 - 0 (disabled) - disable extension. 3392 - 1 (enabled) - enable extension 3393 3394 Default: 0 (disabled) 3395 3396pf_enable - INTEGER 3397 Enable or disable pf (pf is short for potentially failed) state. A value 3398 of pf_retrans > path_max_retrans also disables pf state. That is, one of 3399 both pf_enable and pf_retrans > path_max_retrans can disable pf state. 3400 Since pf_retrans and path_max_retrans can be changed by userspace 3401 application, sometimes user expects to disable pf state by the value of 3402 pf_retrans > path_max_retrans, but occasionally the value of pf_retrans 3403 or path_max_retrans is changed by the user application, this pf state is 3404 enabled. As such, it is necessary to add this to dynamically enable 3405 and disable pf state. See: 3406 https://datatracker.ietf.org/doc/draft-ietf-tsvwg-sctp-failover for 3407 details. 3408 3409 Possible values: 3410 3411 - 1: Enable pf. 3412 - 0: Disable pf. 3413 3414 Default: 1 3415 3416pf_expose - INTEGER 3417 Unset or enable/disable pf (pf is short for potentially failed) state 3418 exposure. Applications can control the exposure of the PF path state 3419 in the SCTP_PEER_ADDR_CHANGE event and access of SCTP_PF-state 3420 transport info via SCTP_GET_PEER_ADDR_INFO sockopt. 3421 3422 Possible values: 3423 3424 - 0: Unset pf state exposure (compatible with old applications). No 3425 event will be sent but the transport info can be queried. 3426 - 1: Disable pf state exposure. No event will be sent and trying to 3427 obtain transport info will return -EACCESS. 3428 - 2: Enable pf state exposure. The event will be sent for a transport 3429 becoming SCTP_PF state and transport info can be obtained. 3430 3431 Default: 0 3432 3433addip_noauth_enable - BOOLEAN 3434 Dynamic Address Reconfiguration (ADD-IP) requires the use of 3435 authentication to protect the operations of adding or removing new 3436 addresses. This requirement is mandated so that unauthorized hosts 3437 would not be able to hijack associations. However, older 3438 implementations may not have implemented this requirement while 3439 allowing the ADD-IP extension. For reasons of interoperability, 3440 we provide this variable to control the enforcement of the 3441 authentication requirement. 3442 3443 == =============================================================== 3444 1 Allow ADD-IP extension to be used without authentication. This 3445 should only be set in a closed environment for interoperability 3446 with older implementations. 3447 3448 0 Enforce the authentication requirement 3449 == =============================================================== 3450 3451 Default: 0 3452 3453auth_enable - BOOLEAN 3454 Enable or disable Authenticated Chunks extension. This extension 3455 provides the ability to send and receive authenticated chunks and is 3456 required for secure operation of Dynamic Address Reconfiguration 3457 (ADD-IP) extension. 3458 3459 Possible values: 3460 3461 - 0 (disabled) - disable extension. 3462 - 1 (enabled) - enable extension 3463 3464 Default: 0 (disabled) 3465 3466prsctp_enable - BOOLEAN 3467 Enable or disable the Partial Reliability extension (RFC3758) which 3468 is used to notify peers that a given DATA should no longer be expected. 3469 3470 Possible values: 3471 3472 - 0 (disabled) - disable extension. 3473 - 1 (enabled) - enable extension 3474 3475 Default: 1 (enabled) 3476 3477max_burst - INTEGER 3478 The limit of the number of new packets that can be initially sent. It 3479 controls how bursty the generated traffic can be. 3480 3481 Default: 4 3482 3483association_max_retrans - INTEGER 3484 Set the maximum number for retransmissions that an association can 3485 attempt deciding that the remote end is unreachable. If this value 3486 is exceeded, the association is terminated. 3487 3488 Default: 10 3489 3490max_init_retransmits - INTEGER 3491 The maximum number of retransmissions of INIT and COOKIE-ECHO chunks 3492 that an association will attempt before declaring the destination 3493 unreachable and terminating. 3494 3495 Default: 8 3496 3497path_max_retrans - INTEGER 3498 The maximum number of retransmissions that will be attempted on a given 3499 path. Once this threshold is exceeded, the path is considered 3500 unreachable, and new traffic will use a different path when the 3501 association is multihomed. 3502 3503 Default: 5 3504 3505pf_retrans - INTEGER 3506 The number of retransmissions that will be attempted on a given path 3507 before traffic is redirected to an alternate transport (should one 3508 exist). Note this is distinct from path_max_retrans, as a path that 3509 passes the pf_retrans threshold can still be used. Its only 3510 deprioritized when a transmission path is selected by the stack. This 3511 setting is primarily used to enable fast failover mechanisms without 3512 having to reduce path_max_retrans to a very low value. See: 3513 http://www.ietf.org/id/draft-nishida-tsvwg-sctp-failover-05.txt 3514 for details. Note also that a value of pf_retrans > path_max_retrans 3515 disables this feature. Since both pf_retrans and path_max_retrans can 3516 be changed by userspace application, a variable pf_enable is used to 3517 disable pf state. 3518 3519 Default: 0 3520 3521ps_retrans - INTEGER 3522 Primary.Switchover.Max.Retrans (PSMR), it's a tunable parameter coming 3523 from section-5 "Primary Path Switchover" in rfc7829. The primary path 3524 will be changed to another active path when the path error counter on 3525 the old primary path exceeds PSMR, so that "the SCTP sender is allowed 3526 to continue data transmission on a new working path even when the old 3527 primary destination address becomes active again". Note this feature 3528 is disabled by initializing 'ps_retrans' per netns as 0xffff by default, 3529 and its value can't be less than 'pf_retrans' when changing by sysctl. 3530 3531 Default: 0xffff 3532 3533rto_initial - INTEGER 3534 The initial round trip timeout value in milliseconds that will be used 3535 in calculating round trip times. This is the initial time interval 3536 for retransmissions. 3537 3538 Default: 3000 3539 3540rto_max - INTEGER 3541 The maximum value (in milliseconds) of the round trip timeout. This 3542 is the largest time interval that can elapse between retransmissions. 3543 3544 Default: 60000 3545 3546rto_min - INTEGER 3547 The minimum value (in milliseconds) of the round trip timeout. This 3548 is the smallest time interval the can elapse between retransmissions. 3549 3550 Default: 1000 3551 3552hb_interval - INTEGER 3553 The interval (in milliseconds) between HEARTBEAT chunks. These chunks 3554 are sent at the specified interval on idle paths to probe the state of 3555 a given path between 2 associations. 3556 3557 Default: 30000 3558 3559sack_timeout - INTEGER 3560 The amount of time (in milliseconds) that the implementation will wait 3561 to send a SACK. 3562 3563 Default: 200 3564 3565valid_cookie_life - INTEGER 3566 The default lifetime of the SCTP cookie (in milliseconds). The cookie 3567 is used during association establishment. 3568 3569 Default: 60000 3570 3571cookie_preserve_enable - BOOLEAN 3572 Enable or disable the ability to extend the lifetime of the SCTP cookie 3573 that is used during the establishment phase of SCTP association 3574 3575 Possible values: 3576 3577 - 0 (disabled) - disable. 3578 - 1 (enabled) - enable cookie lifetime extension. 3579 3580 Default: 1 (enabled) 3581 3582cookie_hmac_alg - STRING 3583 Select the hmac algorithm used when generating the cookie value sent by 3584 a listening sctp socket to a connecting client in the INIT-ACK chunk. 3585 Valid values are: 3586 3587 * sha256 3588 * none 3589 3590 Default: sha256 3591 3592rcvbuf_policy - INTEGER 3593 Determines if the receive buffer is attributed to the socket or to 3594 association. SCTP supports the capability to create multiple 3595 associations on a single socket. When using this capability, it is 3596 possible that a single stalled association that's buffering a lot 3597 of data may block other associations from delivering their data by 3598 consuming all of the receive buffer space. To work around this, 3599 the rcvbuf_policy could be set to attribute the receiver buffer space 3600 to each association instead of the socket. This prevents the described 3601 blocking. 3602 3603 - 1: rcvbuf space is per association 3604 - 0: rcvbuf space is per socket 3605 3606 Default: 0 3607 3608sndbuf_policy - INTEGER 3609 Similar to rcvbuf_policy above, this applies to send buffer space. 3610 3611 - 1: Send buffer is tracked per association 3612 - 0: Send buffer is tracked per socket. 3613 3614 Default: 0 3615 3616sctp_mem - vector of 3 INTEGERs: min, pressure, max 3617 Number of pages allowed for queueing by all SCTP sockets. 3618 3619 * min: Below this number of pages SCTP is not bothered about its 3620 memory usage. When amount of memory allocated by SCTP exceeds 3621 this number, SCTP starts to moderate memory usage. 3622 * pressure: This value was introduced to follow format of tcp_mem. 3623 * max: Maximum number of allowed pages. 3624 3625 Default is calculated at boot time from amount of available memory. 3626 3627sctp_rmem - vector of 3 INTEGERs: min, default, max 3628 Only the first value ("min") is used, "default" and "max" are 3629 ignored. 3630 3631 * min: Minimal size of receive buffer used by SCTP socket. 3632 It is guaranteed to each SCTP socket (but not association) even 3633 under moderate memory pressure. 3634 3635 Default: 4K 3636 3637sctp_wmem - vector of 3 INTEGERs: min, default, max 3638 Only the first value ("min") is used, "default" and "max" are 3639 ignored. 3640 3641 * min: Minimum size of send buffer that can be used by SCTP sockets. 3642 It is guaranteed to each SCTP socket (but not association) even 3643 under moderate memory pressure. 3644 3645 Default: 4K 3646 3647addr_scope_policy - INTEGER 3648 Control IPv4 address scoping (see 3649 https://datatracker.ietf.org/doc/draft-stewart-tsvwg-sctp-ipv4/00/ 3650 for details). 3651 3652 - 0 - Disable IPv4 address scoping 3653 - 1 - Enable IPv4 address scoping 3654 - 2 - Follow draft but allow IPv4 private addresses 3655 - 3 - Follow draft but allow IPv4 link local addresses 3656 3657 Default: 1 3658 3659udp_port - INTEGER 3660 The listening port for the local UDP tunneling sock. Normally it's 3661 using the IANA-assigned UDP port number 9899 (sctp-tunneling). 3662 3663 This UDP sock is used for processing the incoming UDP-encapsulated 3664 SCTP packets (from RFC6951), and shared by all applications in the 3665 same net namespace. This UDP sock will be closed when the value is 3666 set to 0. 3667 3668 The value will also be used to set the src port of the UDP header 3669 for the outgoing UDP-encapsulated SCTP packets. For the dest port, 3670 please refer to 'encap_port' below. 3671 3672 Default: 0 3673 3674encap_port - INTEGER 3675 The default remote UDP encapsulation port. 3676 3677 This value is used to set the dest port of the UDP header for the 3678 outgoing UDP-encapsulated SCTP packets by default. Users can also 3679 change the value for each sock/asoc/transport by using setsockopt. 3680 For further information, please refer to RFC6951. 3681 3682 Note that when connecting to a remote server, the client should set 3683 this to the port that the UDP tunneling sock on the peer server is 3684 listening to and the local UDP tunneling sock on the client also 3685 must be started. On the server, it would get the encap_port from 3686 the incoming packet's source port. 3687 3688 Default: 0 3689 3690plpmtud_probe_interval - INTEGER 3691 The time interval (in milliseconds) for the PLPMTUD probe timer, 3692 which is configured to expire after this period to receive an 3693 acknowledgment to a probe packet. This is also the time interval 3694 between the probes for the current pmtu when the probe search 3695 is done. 3696 3697 PLPMTUD will be disabled when 0 is set, and other values for it 3698 must be >= 5000. 3699 3700 Default: 0 3701 3702reconf_enable - BOOLEAN 3703 Enable or disable extension of Stream Reconfiguration functionality 3704 specified in RFC6525. This extension provides the ability to "reset" 3705 a stream, and it includes the Parameters of "Outgoing/Incoming SSN 3706 Reset", "SSN/TSN Reset" and "Add Outgoing/Incoming Streams". 3707 3708 Possible values: 3709 3710 - 0 (disabled) - Disable extension. 3711 - 1 (enabled) - Enable extension. 3712 3713 Default: 0 (disabled) 3714 3715intl_enable - BOOLEAN 3716 Enable or disable extension of User Message Interleaving functionality 3717 specified in RFC8260. This extension allows the interleaving of user 3718 messages sent on different streams. With this feature enabled, I-DATA 3719 chunk will replace DATA chunk to carry user messages if also supported 3720 by the peer. Note that to use this feature, one needs to set this option 3721 to 1 and also needs to set socket options SCTP_FRAGMENT_INTERLEAVE to 2 3722 and SCTP_INTERLEAVING_SUPPORTED to 1. 3723 3724 Possible values: 3725 3726 - 0 (disabled) - Disable extension. 3727 - 1 (enabled) - Enable extension. 3728 3729 Default: 0 (disabled) 3730 3731ecn_enable - BOOLEAN 3732 Control use of Explicit Congestion Notification (ECN) by SCTP. 3733 Like in TCP, ECN is used only when both ends of the SCTP connection 3734 indicate support for it. This feature is useful in avoiding losses 3735 due to congestion by allowing supporting routers to signal congestion 3736 before having to drop packets. 3737 3738 Possible values: 3739 3740 - 0 (disabled) - Disable ecn. 3741 - 1 (enabled) - Enable ecn. 3742 3743 Default: 1 (enabled) 3744 3745l3mdev_accept - BOOLEAN 3746 Enabling this option allows a "global" bound socket to work 3747 across L3 master domains (e.g., VRFs) with packets capable of 3748 being received regardless of the L3 domain in which they 3749 originated. Only valid when the kernel was compiled with 3750 CONFIG_NET_L3_MASTER_DEV. 3751 3752 Possible values: 3753 3754 - 0 (disabled) 3755 - 1 (enabled) 3756 3757 Default: 1 (enabled) 3758 3759 3760``/proc/sys/net/core/*`` 3761======================== 3762 3763 Please see: Documentation/admin-guide/sysctl/net.rst for descriptions of these entries. 3764 3765 3766``/proc/sys/net/unix/*`` 3767======================== 3768 3769max_dgram_qlen - INTEGER 3770 The maximum length of dgram socket receive queue 3771 3772 Default: 10 3773 3774