| 1 |
# The IKE Scanner (ike-scan) is Copyright (C) 2003-2013 Roy Hills, |
| 2 |
# NTA Monitor Ltd. |
| 3 |
# |
| 4 |
# This file is part of ike-scan. |
| 5 |
# |
| 6 |
# ike-scan is free software: you can redistribute it and/or modify |
| 7 |
# it under the terms of the GNU General Public License as published by |
| 8 |
# the Free Software Foundation, either version 3 of the License, or |
| 9 |
# (at your option) any later version. |
| 10 |
# |
| 11 |
# ike-scan is distributed in the hope that it will be useful, |
| 12 |
# but WITHOUT ANY WARRANTY; without even the implied warranty of |
| 13 |
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
| 14 |
# GNU General Public License for more details. |
| 15 |
# |
| 16 |
# You should have received a copy of the GNU General Public License |
| 17 |
# along with ike-scan. If not, see <http://www.gnu.org/licenses/>. |
| 18 |
# |
| 19 |
# ike-backoff-patterns -- Backoff patterns file for ike-scan |
| 20 |
# |
| 21 |
# Author: Roy Hills <[email protected]> |
| 22 |
# |
| 23 |
# Format: |
| 24 |
# Implementation_Name<Tab>Backoff_Pattern |
| 25 |
# |
| 26 |
# Implementation_Name is a descriptive name for the IKE implementation |
| 27 |
# (and version if applicable). Backoff_Pattern is the observed IKE |
| 28 |
# retransmission backoff pattern for this implementation. |
| 29 |
# |
| 30 |
# The backoff pattern is specified as a comma-separated list of backoff |
| 31 |
# times in seconds. The first number is always zero and represents the first |
| 32 |
# packet received. Subsequent numbers represent the expected delay in |
| 33 |
# seconds after the previous packet. For example, "0, 2, 2, 2" means that a |
| 34 |
# total of four packets are sent with a delay of two seconds between each one. |
| 35 |
# |
| 36 |
# The numbers in the backoff pattern can be decimal numbers e.g. 1.5 for |
| 37 |
# one and a half seconds. You can specify up to a maximum of 6 digits |
| 38 |
# after the decimal point (microsecond resolution), although anything |
| 39 |
# beyond millisecond resolution is not really practical. ike-scan uses |
| 40 |
# a timeval struct to store the backoff pattern entries. |
| 41 |
# |
| 42 |
# You may also specify an per-pattern-entry "fuzz" value in milliseconds |
| 43 |
# by appending /<fuzz> to the backoff time. This will override the default |
| 44 |
# fuzz for that time only. The fuzz value specifies how close the specified |
| 45 |
# time must be to the observed time for a match. E.g. a fuzz of 0 means that |
| 46 |
# the observed timing must match exactly and a fuzz of 1000 means that the |
| 47 |
# times must be within one second of each other (1000ms = 1sec). If no |
| 48 |
# per-pattern-entry fuzz value is specified then a default fuzz value of 100ms |
| 49 |
# (defined by DEFAULT_PATTERN_FUZZ in ike-scan.h) is applied. This default |
| 50 |
# value may be changed with the --fuzz option to ike-scan. |
| 51 |
# |
| 52 |
# Lines beginning with '#' and blank lines are ignored. |
| 53 |
# |
| 54 |
# The input format is quite strict. In particular, the separator between |
| 55 |
# the implementation name and the backoff pattern must be a single TAB and |
| 56 |
# not a space, multiple tabs or spaces, or a mixture of tabs and spaces. |
| 57 |
# |
| 58 |
# If you have problems adding entries, run ike-scan as: |
| 59 |
# ike-scan -v -v -v -o <any-target> |
| 60 |
# To dump the backoff pattern table. |
| 61 |
# |
| 62 |
# There are examples and analysis of the backoff patterns returned by various |
| 63 |
# implementations on the ike-scan wiki at |
| 64 |
# http://www.royhills.co.uk/wiki/index.php/Ike-scan_Documentation |
| 65 |
# |
| 66 |
# You are encouraged to submit comments, improvements or suggestions |
| 67 |
# at the github repository https://github.com/royhills/ike-scan |
| 68 |
# In particular, I would like you to submit any new patterns that you |
| 69 |
# discover. |
| 70 |
# |
| 71 |
|
| 72 |
# Discovered by: Roy Hills, November 2002 |
| 73 |
# Observed on: Cisco 2503 running IP-PLUS-IPsec56 IOS 11.3(11b)T2 |
| 74 |
# Observed on: Cisco 2503 running IP-PLUS-IPsec56 IOS 12.0(28c) |
| 75 |
# Observed on: Cisco PIX 520 running 5.1(2) |
| 76 |
# Observed on: Cisco PIX 520 running 5.2(9) |
| 77 |
# Observed on: Cisco PIX 520 running 5.3(4) |
| 78 |
# Observed on: Cisco PIX 520 running 6.0(4) |
| 79 |
# Observed on: Cisco PIX 520 running 6.1(5) |
| 80 |
# Observed on: Cisco PIX 520 running 6.2(4) |
| 81 |
# Note: Cisco IOS 12.1, 12.2 and 12.3 (and possibly later versions as well) |
| 82 |
# have a different pattern: 0,10,10,10,10,10. This is listed later |
| 83 |
# under a different name. |
| 84 |
# Note: PIX OS 6.3 has a different pattern: 0, 5, 5, 5, 5, 5. This is listed |
| 85 |
# under a different name. |
| 86 |
# Note: IPsec was introduced in Cisco IOS 11.3. Previous versions only |
| 87 |
# supported the Cisco proprietary CET encryption. |
| 88 |
Cisco IOS 11.3 or 12.0 / PIX <= 6.2 0, 15, 15 |
| 89 |
|
| 90 |
# Discovered by: Tony Lloyd, May 2006 |
| 91 |
# Observed on: Cisco PIX 520 running PIXOS 6.3(5) |
| 92 |
Cisco PIX 6.3 0, 5, 5, 5, 5, 5 |
| 93 |
|
| 94 |
# 1st Pattern Discovered by: Roy Hills, November 2002 |
| 95 |
# Observed on: Cisco VPN Concentrator 3005 running (unknown version) |
| 96 |
# Observed on: VPN 3000 Concentrator Version 3.6.3.Rel Oct 04 2002 16:23:00 |
| 97 |
# Observed on: VPN 3005 Concentrator Version 4.7.2.L Jun 11 2007 14:07:29 |
| 98 |
# 2nd Pattern Discovered by: Guy Widloecher, March 2003 |
| 99 |
# Observed on: Cisco VPN 3005 Concentrator Version 3.5.2 |
| 100 |
# This second pattern may be due to the earlier software version, or it could |
| 101 |
# be a missed packet. |
| 102 |
# There are several models: 3005, 3015, 3020, 3030, 3060 and 3080, which are |
| 103 |
# equivalent to the old Altiga C5, C15 Etc. All models are believed to share |
| 104 |
# the same backoff pattern. |
| 105 |
# The first pattern is also generated by Cisco PIX 7.0 |
| 106 |
# Observed on: Cisco PIX 515E running PIX OS 7.0(7) |
| 107 |
Cisco VPN Concentrator or PIX 7.0 0, 8, 8, 8 |
| 108 |
Cisco VPN Concentrator 0, 8, 8 |
| 109 |
|
| 110 |
# Discovered by: Roy Hills, December 2002 |
| 111 |
# Observed on: Checkpoint Firewall-1 4.0 SP7 on Windows NT Workstation 4.0 SP6a |
| 112 |
Firewall-1 4.0 0, 3, 3, 3, 3, 6, 6, 6, 6, 6, 6, 6 |
| 113 |
|
| 114 |
# Discovered by: Roy Hills, November 2002 |
| 115 |
# Observed on: Checkpoint Firewall-1 4.1 SP6 on Windows NT Server 4.0 SP6a |
| 116 |
# Observed on: Checkpoint Firewall-1 NG base on Windows NT Server 4.0 SP6a |
| 117 |
# Observed on: Checkpoint Firewall-1 NG FP2 on Windows NT Server 4.0 SP6a |
| 118 |
# Observed on: Checkpoint Firewall-1 NGX R60 on Windows 2003 Server |
| 119 |
Firewall-1 4.1/NG/NGX 0, 2, 2, 2, 2, 2, 2, 4, 4, 4, 4, 4 |
| 120 |
|
| 121 |
# Discovered by: Roy Hills, December 2002 |
| 122 |
# Observed on: FreeS/WAN 1.9 on Debian Linux 2.2r7 (Potato) with 2.2.17 Kernel |
| 123 |
# Observed on: strongSwan 4.0.5 on Debian Etch with 2.6.18 kernel |
| 124 |
# Observed on: OpenSwan 2.2.0 on Debian Sarge with 2.6.16 kernel |
| 125 |
Linux FreeS/WAN, OpenSwan, strongSwan 0, 10, 20 |
| 126 |
|
| 127 |
# Discovered by: Roy Hills, November 2002 |
| 128 |
# Additional fuzz on 1st retry suggested by Florent Trupheme, April 2005 |
| 129 |
# Observed on: Nortel Contivity 2500, OS version V4.06-120 |
| 130 |
# Observed on: Nortel Contivity 1600, OS version V3.60-45 |
| 131 |
Nortel Contivity 0, 16/600, 16, 16 |
| 132 |
|
| 133 |
# Discovered by: Roy Hills, December 2002 |
| 134 |
# Observed on: Watchguard Firebox 700 v6.1 |
| 135 |
# Observed on: Gnat Box (version unknown) |
| 136 |
# Observed on: Cisco 2503 running IP-PLUS-IPsec56 IOS 12.1(27a) |
| 137 |
# Observed on: Cisco 2503 running IP-PLUS-IPsec56 IOS 12.2(29) |
| 138 |
# Observed on: Cisco 2621 running IP/FW/IDS Plus IPsec 3DES Basic 12.3(17a) |
| 139 |
# Observed on: Cisco 1721 running Advanced Security 12.4(13a) |
| 140 |
# Observed on: Cisco 7206 running Advanced Security 15.0(1)M |
| 141 |
# Observed on: racoon 0.6.6 running on Debian Etch |
| 142 |
# Note that the Gnat box has a much larger variance than the watchguard, but |
| 143 |
# they are both essentially the same pattern. |
| 144 |
Cisco IOS 12.1, 12.2, 12.3, 12.4 or 15.0 / Watchguard Firebox / Gnat Box / racoon 0, 10/1000, 10/1000, 10/1000, 10/1000, 10/1000 |
| 145 |
|
| 146 |
# Discovered by: Roy Hills, December 2002 |
| 147 |
# Observed on: Windows 2000 Server SP1 |
| 148 |
# Observed on: Windows XP Pro SP1 |
| 149 |
# Note: Backoff fingerprinting cannot distinguish between 2000, 2003 and XP, |
| 150 |
# but vendor IDs can. |
| 151 |
Windows 2000, 2003 or XP 0, 1, 2, 4, 8, 16, 32 |
| 152 |
|
| 153 |
# 1st pattern Discovered by: Thomas Walpuski, Jan 2003 |
| 154 |
# Observed on: Various OpenBSD systems running isakmpd |
| 155 |
# 2nd pattern discovered by: Marco Ivaldi <[email protected]>, Jan 2003 |
| 156 |
# Observed on: OpenBSD 3.2 |
| 157 |
# Observed on: FreeBSD 4.7-Stable with isakmpd-20021118 and OpenBSD 3.1 |
| 158 |
# Note: OpenBSD isakmpd is highly configurable, so it's difficult to get one |
| 159 |
# pattern which will match all possible backoff pattern. |
| 160 |
# Hakan Olsson has informed me that the actual algorithm used by isakmpd |
| 161 |
# is "5 + 2*<retrans#>" which can be found in transport.c, ca line 310. |
| 162 |
# Both patterns match this algorithm: the first with the default |
| 163 |
# retransmission limit of 3 and the second with retransmits set to 5. |
| 164 |
# It has also been pointed out that isakmpd can run on many platforms |
| 165 |
# other than FreeBSD and OpenBSD, so the inclusion of these OS names in |
| 166 |
# the pattern are misleading. However, I'm leaving the names unchanged |
| 167 |
# in case someone relies on them in the program output. |
| 168 |
FreeBSD/OpenBSD-isakmpd 0, 7, 9, 11 |
| 169 |
FreeBSD/OpenBSD-isakmpd 0, 7, 9, 11, 13, 15 |
| 170 |
|
| 171 |
# Discovered by: Paul van Maaren, January 2003 |
| 172 |
# 1st pattern observed Jan 2003 on: FreeBSD-4.7 STABLE with racoon-20021120a |
| 173 |
# 2nd pattern observed Jun 2005 on: FreeBSD-5.3-RELEASE with racoon-20050510a |
| 174 |
KAME/racoon-1 0, 20, 20, 20, 20, 20 |
| 175 |
KAME/racoon-2 0, 21.5, 21.5, 21.5, 21.5, 21.5 |
| 176 |
|
| 177 |
# Discovered by: Iain Lewis, February 2003 |
| 178 |
# Observed on: Netscreen 500 |
| 179 |
# Observed on: Netscreen 5GT running ScreenOS 5.1.0r1.0 |
| 180 |
Juniper-Netscreen 0, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4 |
| 181 |
|
| 182 |
# Discovered by: Doug Monroe, January 2003 |
| 183 |
# Observed on: Watchguard SOHO v5.1.7 |
| 184 |
# Note: There is quite a bit of variation with this pattern, hence the |
| 185 |
# per-entry fuzz specifications. |
| 186 |
watchguard-soho 0, 1/500, 9.5/3000, 1.5/2000, 9.5/3000, 1.5/2000, 9.5/3000, 1.5/2000, 9.5/3000, 1.5/2000, 9.5/3000, 1.5/2000 |
| 187 |
|
| 188 |
# Discovered by: Christopher Harrington, February 2003 |
| 189 |
# Observed on: SonicWall Pro 200 |
| 190 |
# Note: There is quite a bit of variation with this pattern, hence the |
| 191 |
# per-entry fuzz specifications. |
| 192 |
sonicwall-pro 0, 6.5/1500, 9/1000, 13/2000, 22/3000 |
| 193 |
|
| 194 |
# Discovered by: Florent Trupheme, April 2005 |
| 195 |
# Observed on Side Winder G2 Firewall |
| 196 |
# Also observed on a separate Sidewinder G2 by Tim Ecott, July 2005 |
| 197 |
# Two different patterns have been observed: the 0,3,6,15,48 appears to be the |
| 198 |
# more common one, but 0,3,4,12,36 has also been seen. |
| 199 |
SideWinder G2 Firewall 0, 3, 6, 15, 48 |
| 200 |
SideWinder G2 Firewall 0, 3, 4, 12, 36 |
| 201 |
|
| 202 |
# Discovered by: Florent Trupheme, April 2005 |
| 203 |
# Observed on SonicWall unknown version |
| 204 |
# Interestingly, this is different from the earlier sonicwall-pro entry. |
| 205 |
Sonic Wall 0, 5, 8, 18 |
| 206 |
|
| 207 |
# Discovered by: RealRancor, September 2013 |
| 208 |
# DELL SonicWALL appliance (version unknown) |
| 209 |
# There's quite a bit of variation in the timings, hence the per-entry fuzz |
| 210 |
# settings. |
| 211 |
# This may be the same patterm as "Sonic Wall" above, but we don't have |
| 212 |
# enough data points to be sure. |
| 213 |
DELL SonicWall 0, 5/1000, 10/2000, 20/3000 |
| 214 |
|
| 215 |
# Discovered by: Roy Hills, December 2003 |
| 216 |
# Observed on: Windows 2003 Server Enterprise Edition, Intel platform |
| 217 |
# Note: The 2nd packet delay has been observed to vary between 0.5 and 1.5 sec. |
| 218 |
# Otherwise, this is the same pattern as Windows 2000. Note that if the |
| 219 |
# Win-2003 server happens to pick a delay of around 1 sec for the 2nd |
| 220 |
# Packet, then ike-scan will mis-identify as Win-2000. Vendor ID |
| 221 |
# payloads can distinguish Win-2003 from Win-2000 in any event, but |
| 222 |
# that's another story. |
| 223 |
Windows-2003 0, 1/600, 2, 4, 8, 16, 32 |
| 224 |
|
| 225 |
# Discovered by: Bob Davies, March 2004 |
| 226 |
# Observed on Lynksys router, Unknown version |
| 227 |
Lynksys 0, 15/500, 15, 15 |
| 228 |
|
| 229 |
# Discovered by: Tony Lloyd, January 2005 |
| 230 |
# Observed on: suspected bordermanager box |
| 231 |
# July 2006: Observed on: BorderManager 3.8 on NetWare 6.5 |
| 232 |
Novell-BorderManager 0, 4.5/1000, 6.9, 9.8 |
| 233 |
|
| 234 |
# Discovered by: Tony Lloyd, April 2005 |
| 235 |
# Observed on: Draytek ADSL Routers (several different versions) |
| 236 |
draytek 0, 3, 6 |
| 237 |
|
| 238 |
# Discovered by: Tony Lloyd, June 2005 |
| 239 |
# Observed on: Cyberguard Firewall (exact model and version unknown) |
| 240 |
cyberguard 0, 3.5, 7, 10, 10 |
| 241 |
|
| 242 |
# Discovered by: Tony Lloyd, June 2005 |
| 243 |
# Observed on: Fortinet FortiGate Firewall (exact model and version unknown) |
| 244 |
# Note: some FortiGate Firewalls appear to have a 0,10,20 pattern instead |
| 245 |
FortiGate 0, 6, 12 |
| 246 |
|
| 247 |
# Discovered by: Tony Lloyd, August 2005 |
| 248 |
# Observed on: Avaya VSU 100R |
| 249 |
# The backoffs have quite a bit of variance, but the delays always seem to be |
| 250 |
# between 13.3 and 15.7 seconds. |
| 251 |
Avaya VSU 0, 14.5/1200, 14.5/1200, 14.5/1200, 14.5/1200 |
| 252 |
|
| 253 |
# Discovered by: Tony Lloyd, September 2005 |
| 254 |
# Observed on: Stonegate V2.2.0 on Linux |
| 255 |
Stonegate 0, 0.5, 1, 2, 4, 8, 16, 30, 30, 30, 30 |
| 256 |
|
| 257 |
# Discovered by Paul Askew, December 2005 |
| 258 |
# Observed on: Netgear FVS328 ProSafe VPN Firewall Version 1.0.15 |
| 259 |
# The time between the first and second packets has quite a bit of variance, |
| 260 |
# but the delays between subsequent packets do not. |
| 261 |
Netgear ProSafe 0, 4.5/1000, 5, 5, 5 |
| 262 |
|
| 263 |
# Discovered by Paul Askew, December 2005 |
| 264 |
# Observed on: Netgear DG834V2 ADSL Firewall Router Version 2.10.22 |
| 265 |
# Interestingly, the backoff pattern for this device differs from that for |
| 266 |
# the Netgear FVS328 ProSafe VPN Firewall. |
| 267 |
Netgear ADSL Firewall Router 0, 10, 20 |
| 268 |
|
| 269 |
# Discovered by Roy Hills, October 2006 |
| 270 |
# Observed on: Linksys Etherfast DSL/Cable VPN Router, Model BEFVP41 V2 s/w 1.01.04 |
| 271 |
# Not really a pattern as the Linksys doesn't retransmit at all. However, this |
| 272 |
# lack of retransmission is sufficiently unusual to be a pattern itself. |
| 273 |
Linksys Etherfast 0 |
| 274 |
|
| 275 |
# Discovered by Roy Hills, November 2005 |
| 276 |
# Observed on: IBM RS/6000 running AIX 5L 5.3 |
| 277 |
# The AIX backoff is unusual because it goes on for a long time (about 19 |
| 278 |
# minutes) and it has a hiccup at one point where it sends two packets with |
| 279 |
# the same backoff. With the standard ike-scan backoff timeout, you will only |
| 280 |
# see the first three packets. Longer timeouts let you see more packets. |
| 281 |
# I have added multiple patterns to allow matching with any backoff timeout |
| 282 |
# above 32 seconds. |
| 283 |
IBM AIX 0, 16, 32 |
| 284 |
IBM AIX 0, 16, 32, 64 |
| 285 |
IBM AIX 0, 16, 32, 64, 128, 128 |
| 286 |
IBM AIX 0, 16, 32, 64, 128, 128, 256 |
| 287 |
IBM AIX 0, 16, 32, 64, 128, 128, 256, 512 |
| 288 |
|
| 289 |
# Discovered by Roy Hills, January 2007 |
| 290 |
# Observed on: Solaris 9/SPARC running on Ultra-5 |
| 291 |
# Observed on: Solaris 10/Intel running on VMware |
| 292 |
# |
| 293 |
# Sun's IPsec was introduced in Solaris 8, but this was manual keying only; |
| 294 |
# IKE keying was not introduced until Solaris 9. |
| 295 |
Sun Solaris 0, 0.5, 1, 2, 4, 8 |