|
LOGS: GIAC GCIA Version 3.4 Practical Detect (Dana Webber): msg#00005security.intrusions
I posted a rough draft of this detect a few months ago but got no feedback. I'm handing in the practical in a few days. Here is the final version. The old rough draft is at http://cert.uni-stuttgart.de/archive/intrusions/2004/01/msg00133.html Thanks a zillion! ============================================================================== GIAC GCIA Version 3.4 Practical Detect Possible VMware penetration test. Source of Trace: http://www.incidents.org/logs/2003.12.15.tgz Detect was generated by: Detect File Overview Archive URL http://www.incidents.org/logs/2003.12.15.tgz files in archive 2003.12.15,.1 2003.12.15.2, ..., 2003.12.15.14 File Format libpcap, Snaplen = 96 bytes Start time 11/18/2003 13:57:23.130647 End time 11/18/2003 15:15:57.147884 Duration 1 hour and 18 minutes. Packets captured 449147 total Alerts 31791 distinct alerts 66 Source MAC's 30 Destination MAC's 39 Alerts to 0:50:56:40:0:6d 28633 MAC address obfuscation is normally not done and IP address obfuscation was not necessary because all the detected IP's were reserved for private networks None of the IP's are legal on the Internet. They were all in reserved ranges. Command: whois $ip @whois.arin.net; NetRange: 10.0.0.0 - 10.255.255.255 This block is reserved for special purposes. NetRange: 192.168.0.0 - 192.168.255.255 This block is reserved for special purposes. NetRange: 172.16.0.0 - 172.31.255.255 This block is reserved for special purposes. NetRange: 224.0.0.0 - 239.255.255.255 This block is reserved for special purposes. It appears that every packet was captured in libpcap format with snaplen 96. It was assumed that all traffic at the target's LAN connection was captured. This may not be correct. Most of the systems were not analyzed. It is assumed they were monitoring the test. Eight consecutive Cisco MAC's were detected but none were associated with any IP's. They were transmitting data link level spanning tree data. This could indicate a large router that was not allowed to transmit IP traffic. It may have been used for the data capture. Probability the source address was spoofed: 50% The places I have worked at would buy a Smartbits [1] and connect it with one Ethernet cable to the unit to be tested and then "run a test". The Smartbits would generate the traffic and capture the data by itself. Other companies make similar devices. Program that was used to generate the alert file.: #!/bin/sh #A problem with Snort-2.0.5 is that if a packet triggers two alerts then the second #alert takes precedence. This produced ICMP alerts with "(Undefined Code!)". #The fix:move alerts with "(Undefined Code!)" in file icmp-info.rules to the top. cd /1/backup/giac/gcia_assignment COUNTER='1'; while [ "$COUNTER" != 15 ]; do RAWPATH="incidents.org.logs.raw.2003.12.15/2003.12.15.$COUNTER"; echo $RAWPATH ls -l $RAWPATH SNORTPATH="snort/snort-2.0.5-bin"; LOGPATH="p2_detect_1/logs"; CONFPATH="snort/snort.conf"; COMMAND="$SNORTPATH -X -d -e -c $CONFPATH -r $RAWPATH -l $LOGPATH" echo "$COMMAND"; `$COMMAND` COUNTER=$[$COUNTER + 1]; done; #command: snort -X -d -e -c $SNORT_CONF -r $RAW_FILE -l $LOG_FILE #-X means dump the raw packet data starting at the link layer #-d means dump the Application Layer #-e means display the second layer header info #$RAW_FILE is varied from 2003.12.15.1 to 2003.12.15.14 2.1.3. Description of attack: One out of every 27 packets triggered an alert. Most MAC's had only one or two IP's associated with them. Usually a LAN has one or more routers and each router has a MAC associated with many IP's. There were 18 distinct source IP's and 1542 distinct destination IP's detected for MAC 0:50:56:40:0:6d . However, ninety percent of the alerts have 0:50:56:40:0:6d as the destination MAC and the three top alert source MAC's were registered to Apple Computer, Intel and Dell. Furthermore the MAC 0:50:56:xx.xx.xx is registered to VMware INC which does not make routers. VMware enables multiple virtual machines on one physical system. Each virtual machine needs a distinct IP. This may be the reason that the this MAC had 18 IP's. Normally most of the attacks all have the same source MAC because they come from the Internet and go through a router/gateway/firewall. A University campus link would likely have attacks in both directions. The only reasonable explanation here is a penetration test of a VMware system and that all the attacks are part of the same test. Such a test set-up should be isolated therefore an Internet gateway is not required. Manufacturers of equipment on the LAN. Ando makes expensive test equipment. [2] Ambit makes expensive test equipment.[3] Abocom is a OEM manufacturer [4]. VMware, the MAC was probably programmed in specifically for testing.[5] Compaq, makes a variety of PC products. Gateway, makes a variety of PC products. Dell, makes a variety of PC products. 3Com, makes NIC's and Ethernet switches Intel makes a variety of PC products. Connectix made some video conferencing equipment, Microsoft bought them.[6] Sony makes products in every category.[7] DOD MAC's are sometimes used by companies that do have not registered a OUI.[8] Commands used to generate OUI Registrant [9], MAC, and IP list Command used to obtain the list of source MAC addresses. j=1;while [ "$j" != 15 ];do /usr/sbin/tcpdump -ennr ../incidents.org.logs.raw.2003.12.15/2003.12.15.$j \ |cut -d\ -f 2|grep ':'|sort|uniq;j=$[$j+1];done|sort|uniq Command used to obtain the list of destination MAC addresses. j=1;while [ "$j" != 15 ];do /usr/sbin/tcpdump -ennr ../incidents.org.logs.raw.2003.12.15/2003.12.15.$j |\ cut -d\ -f 3|grep ':'|sort|uniq;j=$[$j+1];done|sort|uniq Commandsused to obtain the list of source MAC and IP pairs. j=1;while [ "$j" != 15 ];do /usr/sbin/tcpdump -nne ip - \ incidents.org.logs.raw.2003.12.15/2003.12.15.$j >> tcpdump.eth.5.txt;j=$[$j+1];done cut -d\ -f2,6 tcpdump.eth.5.txt|sed 's/^\(.* .*\..*\..*\..*\)\..*$/\1/' |sort|uniq Commands used to obtain the list of destination MAC and IP pairs. j=1;while [ "$j" != 15 ];do /usr/sbin/tcpdump -nne ip -r\ ../incidents.org.logs.raw.2003.12.15/2003.12.15.$j >> tcpdump.eth.5.txt;j=$[$j+1];done; cut -d\ -f3,8 tcpdump.eth.5.txt|sed 's/^\(.* .*\..*\..*\..*\)\..*$/\1/' |sort|uniq OUI Registrant [8], MAC, and IP list Point Multimedia Systems 0:0:39:f2:67:88 10.10.10.117 Ando Electric Corporation 0:0:e2:92:ee:f 0:0:e2:94:b0:2a 10.10.10.222 10.10.10.226 3COM CORPORATION 0:1:2:79:91:ed 0:1:3:88:29:92 0:4:76:45:61:39 10.10.10.112 10.10.10.234 10.10.10.195 Compaq Computer Co 0:2:a5:b6:e2:e3 10.10.10.186 Intel Corporation 0:a0:c9:ba:6d:85 0:3:47:8c:89:c2 10.10.10.196 10.10.10.165 192.168.117.1 192.168.213.1 Connectix 0:3:ff:df:95:84 10.10.10.228 VMWare, Inc. 0:50:56:40:0:64 0:c:29:14:1e:63 0:c:29:39:6e:67 0:c:29:9e:ef:53 0:50:56:40:0:6d 10.10.10.2 10.10.10.142 10.10.10.160 10.10.10.224 10.10.10.1, 10.30.30.2, 172.20.11, 172.20.11.2, 172.20.11.3, 172.20.11.52, 172.20.11.80, 172.20.201.1 172.20.201.135, 172.20.201.198, 172.20.201.2, 192.168.17.129, 192.168.17.135, 192.168.17.2, 192.168.17.65, 192.168.17.66, 192.168.17.68, 192.168.22.207 Dell Computer Corp. 0:6:5b:d8:bf:ed 0:6:5b:e6:f8:43 0:8:74:5:b7:f8 0:8:74:7:31:ee 10.10.10.122 10.10.10.231 10.10.10.147 10.10.10.111, 172.16.8.189 Dell ESG PCBA Test 0:b:db:17:f4:c9 0:b:db:9b:46:fe 0:b:db:df:53:8d 10.10.10.194, 169.254.135.50, 172.16.9.13, 192.168.222.1 192.168.84.1 10.10.10.164 10.10.10.123 IBM Corporation 0:9:6b:2:e9:3d 10.10.10.212, 172.16.8.229 Apple Computer, Inc. 0:a:95:7c:24:0 0:a:95:d9:95:84 10.10.10.113 10.10.10.232 AMBIT MICROSYSTEMS CORP 0:d0:59:c6:5e:14 10.10.10.141 10.10.10.144 238.122.10.140 Cisco Systems 0:d:bc:17:4:ce N/A 0:d:bc:17:4:cf N/A 0:d:bc:17:4:d0 N/A 0:d:bc:17:4:d2 N/A 0:d:bc:17:4:d4 N/A 0:d:bc:17:4:d5 N/A 0:d:bc:17:4:d6 N/A 0:d:bc:17:4:d8 N/A AboCom Systems, Inc 0:e0:98:a1:7f:da 10.10.10.174 GATEWAY 2000 0:e0:b8:3d:20:40 10.10.10.214 SONY CORPORATION LTD. 8:0:46:79:f7:7c 10.10.10.230 DoD Internet Multicast (RFC-1112) 1:0:5e:0:0:16 224.0.0.22 DoD Internet Multicast (RFC-1112) 1:0:5e:0:0:2 224.0.0.2 DoD Internet Multicast (RFC-1112) 1:0:5e:0:0:5 224.0.0.5 DoD Internet Multicast (RFC-1112) 1:0:5e:0:0:6 224.0.0.6 DoD Internet Multicast (RFC-1112) 1:0:5e:37:96:d0 229.55.150.208 DoD Internet Multicast (RFC-1112) 1:0:5e:7a:a:8c 238.122.10.140 DoD Internet Multicast (RFC-1112) 1:0:5e:7f:ff:fa 239.255.255.250 DoD Internet Multicast (RFC-1112) 1:0:5e:7f:ff:fd 239.255.255.253 DoD Internet Multicast (RFC-1112) Cisco Inter-Switch Link 1:0:c:0:0:0 Cisco Discovery Protocol 1:0:c:cc:cc:cc Spanning tree Multicast 1:80:c2:0:0:0 There were many suspicious alerts. They are sorted in order of occurrences. command: grep "\[\*\*\] .* \[\*\*\]" alert | sort | uniq -c | sort -nr 18176 [**] [111:9:1] (spp_stream4) STEALTH ACTIVITY (NULL scan) detection [**] 5041 [**] [1:465:1] ICMP ISS Pinger [**] 2144 [**] [1:620:3] SCAN Proxy (8080) attempt [**] 1326 [**] [1:402:4] ICMP Destination Unreachable (Port Unreachable) [**] 181 [**] [116:97:1] (snort_decoder): Short UDP packet, length field > payload length 1020 [**] [1:474:1] ICMP superscan echo [**] 690 [**] [1:399:4] ICMP Destination Unreachable (Host Unreachable) [**] 418 [**] [1:401:4] ICMP Destination Unreachable (Network Unreachable) [**] 394 [**] [1:615:4] SCAN SOCKS Proxy attempt [**] 287 [**] [1:1417:2] SNMP request udp [**] 146 [**] [1:1420:2] SNMP trap tcp [**] 140 [**] [1:1418:2] SNMP request tcp [**] 116 [**] [1:1421:2] SNMP AgentX/tcp request [**] 105 [**] [1:618:4] SCAN Squid Proxy attempt [**] 70 [**] [1:1443:2] TFTP GET passwd [**] 60 [**] [1:1411:3] SNMP public access udp [**] 59 [**] [1:467:1] ICMP Nemesis v1.1 Echo [**] 46 [**] [1:1419:2] SNMP trap udp [**] 45 [**] [1:408:4] ICMP Echo Reply [**] 29 [**] [1:553:4] POLICY FTP anonymous login attempt [**] 28 [**] [1:365:5] ICMP PING (Undefined Code!) [**] 27 [**] [1:628:2] SCAN nmap TCP [**] 24 [**] [111:10:1] (spp_stream4) STEALTH ACTIVITY (XMAS scan) detection [**] 21 [**] [111:12:1] (spp_stream4) NMAP FINGERPRINT (stateful) detection [**] 18 [**] [1:449:4] ICMP Time-To-Live Exceeded in Transit [**] 14 [**] [1:524:6] BAD-TRAFFIC tcp port 0 traffic [**] 14 [**] [1:361:7] FTP site exec [**] 12 [**] [1:2049:1] MS-SQL ping attempt [**] 10 [**] [1:491:6] INFO FTP Bad login [**] 10 [**] [1:1413:2] SNMP private access udp [**] 8 [**] [1:1777:2] FTP EXPLOIT STAT * dos attempt [**] 6 [**] [1:566:3] POLICY PCAnywhere server response [**] 6 [**] [1:528:4] BAD-TRAFFIC loopback traffic [**] 6 [**] [1:1957:3] RPC sadmind UDP PING [**] 6 [**] [1:1893:1] SNMP missing community string attempt [**] 6 [**] [1:1867:1] MISC xdmcp info query [**] 6 [**] [1:1504:5] MISC AFS access [**] 6 [**] [105:1:1] spp_bo: Back Orifice Traffic detected (key: 31337) [**] 4 [**] [1:453:4] ICMP Timestamp Request [**] 4 [**] [1:451:4] ICMP Timestamp Reply [**] 4 [**] [1:388:4] ICMP Address Mask Request [**] 4 [**] [1:382:4] ICMP PING Windows [**] 4 [**] [1:336:5] FTP CWD ~root attempt [**] 4 [**] [1:237:1] DDOS Trin00:MastertoDaemon(defaultpassdetected!) [**] 4 [**] [1:1992:1] FTP LIST directory traversal attempt [**] 4 [**] [1:1449:3] POLICY FTP anonymous (ftp) login attempt [**] 3 [**] [1:604:5] RSERVICES rsh froot [**] 3 [**] [1:501:2] MISC source route lssre [**] 3 [**] [1:1432:4] P2P GNUTella GET [**] 2 [**] [1:718:6] TELNET login incorrect [**] 2 [**] [1:659:4] SMTP expn decode [**] 2 [**] [1:356:5] FTP passwd retrieval attempt [**] 2 [**] [1:332:5] FINGER 0 query [**] 2 [**] [1:330:6] FINGER redirection attempt [**] 2 [**] [1:327:5] FINGER remote command pipe execution attempt [**] 2 [**] [1:326:5] FINGER remote command ; execution attempt [**] 2 [**] [1:323:4] FINGER root query [**] 2 [**] [1:1728:2] FTP CWD ~<CR><NEWLINE> attempt [**] 1 [**] [1:489:5] INFO FTP No Password [**] 1 [**] [1:404:4] ICMP Destination Unreachable (Protocol Unreachable) [**] 1 [**] [1:335:4] FTP .rhosts [**] 1 [**] [1:255:8] DNS zone transfer TCP [**] 1 [**] [1:245:1] DDOS mstream handler ping to agent [**] 1 [**] [1:239:1] DDOS shaft handler to agent [**] 1 [**] [1:236:3] DDOS Stacheldraht client check gag [**] 1 [**] [1:1928:3] FTP shadow retrieval attempt [**] 1 [**] [1:1919:3] FTP CWD overflow attempt [**] 1 [**] [1:1444:2] TFTP Get [**] Sorting by source IP shows the top two alert sources. Otherwise it is confusing list. command: egrep "IpLen" alert|cut -d \ -f 1|cut -d ":" -f 1|sort|uniq -c|sort -nr 18202 10.10.10.113 Apple Computer, Inc 9211 10.10.10.165 Intel Corporation 912 10.10.10.164 Dell ESG PCBA Test 711 172.20.201.2 VMWare, Inc 669 10.10.10.231 Dell Computer Corp 650 10.10.10.224 VMWare, Inc. 573 10.30.30.2 VMWare, Inc 433 10.10.10.234 3COM CORPORATION 425 10.10.10.1 VMWare, Inc 390 10.10.10.141 AMBIT MICROSYSTEMS CORP 378 10.10.10.2 VMWare, Inc. 236 10.10.10.195 3COM CORPORATION 230 172.20.201.198 VMWare, Inc. 198 172.20.201.135 VMWare, Inc. 196 172.20.201.1 VMWare, Inc. 140 10.10.10.212 IBM Corporation 132 10.10.10.112 3COM CORPORATION 129 192.168.17.2 VMWare, Inc. 80 10.10.10.194 Dell ESG PCBA Test 57 172.20.11.3 VMWare, Inc 46 10.10.10.186 Compaq Computer Corporation 41 10.10.10.174 AboCom Systems, Inc 34 10.10.10.196 Intel Corporation 23 10.10.10.228 Connectix 20 10.10.10.232 Apple Computer, Inc. 19 172.20.11.2 VMWare, Inc. 15 0.0.0.0 <Used For Bootp> 12 10.10.10.226 Ando Electric Corporation 11 10.10.10.230 SONY CORPORATION LTD. 9 10.10.10.222 Ando Electric Corporatio 7 10.10.10.160 VMWare, Inc. 7 10.10.10.142 VMWare, Inc. 7 10.10.10.122 Ando Electric Corporation 6 169.254.135.50 Dell ESG PCBA Test 4 192.168.17.68 VMWare, Inc. 3 238.122.10.140 AMBIT MICROSYSTEMS CORP 2 172.20.11.80 VMWare, Inc. 2 172.20.11.52 VMWare, Inc. 2 10.10.10.214 GATEWAY 2000 1 172.20.11.1 VMWare, Inc. 1 10.10.10.147 Dell Computer Corp. 1 10.10.10.144 AMBIT MICROSYSTEMS CORP 1 10.10.10.111 Dell Computer Corp. Sort by destination IP: VMWare was the target. Here are the top 42 destination. command: grep "IpLen" alert|cut -d \ -f 3|cut -d ":" -f 1|sort|uniq -c|sort -nr|head -n 42 6650 192.168.17.129 VMWare, Inc 6546 192.168.17.68 VMWare, Inc 5321 192.168.17.135 VMWare, Inc 1375 10.10.10.165 Intel Corporation 994 172.20.201.2 VMWare, Inc 876 172.20.201.198 VMWare, Inc 795 172.20.201.135 VMWare, Inc 780 172.20.201.1 VMWare, Inc 581 10.10.10.224 VMWare, Inc 471 172.20.201.3 VMWare, Inc 407 172.20.11.2 VMWare, Inc 351 10.10.10.164 Dell ESG PCBA Tes 293 192.168.17.66 VMWare, Inc 238 192.168.17.67 VMWare, Inc 235 149.134.30.62 VMWare, Inc 189 149.134.52.149 VMWare, Inc 176 172.22.201.1 VMWare, Inc 156 172.20.11.80 VMWare, Inc. 122 172.22.201.2 VMWare, Inc 120 192.168.22.207 VMWare, Inc. 120 192.168.17.1 Intel Corporation 113 172.20.11.3 VMWare, Inc. 111 10.10.10.255 BROADCAST ADDRESS 105 10.10.10.212 IBM Corporation 91 10.10.10.195 3COM CORPORATION 89 10.10.10.226 Ando Electric Corporation 80 10.10.10.231 Dell Computer Corp. 64 172.22.201.3 VMWare, Inc 56 255.255.255.255 BROADCAST ADDRESS 51 172.20.11.52 VMWare, Inc 39 10.10.10.142 VMWare, Inc 37 10.10.10.222 Ando Electric Corporation 36 172.11.11.80 VMWare, Inc 36 10.10.10.112 3COM CORPORATION 34 10.10.10.2 VMWare, Inc 24 229.55.150.208 DoD Internet Multicast (RFC-1112) 24 172.20.201.0 INVALID 24 10.10.10.234 3COM CORPORATION 23 192.168.17.65 VMWare, Inc. 22 10.10.10.141 AMBIT MICROSYSTEMS CORP 21 172.10.11.80 VMWare, Inc 20 10.10.10.122 Dell Computer Corp Sorting by destination MAC shows that the attacks targeted one particular system. command: grep " -> " alert|grep -v IpLen |cut -d \ -f 4|sort|uniq -c|sort -nr 28633 0:50:56:40:0:6D VMWare, Inc 1375 0:3:47:8C:89:C2 Intel Corporation 581 0:C:29:9E:EF:53 VMWare, Inc 351 0:B:DB:9B:46:FE Dell ESG PCBA Test 173 FF:FF:FF:FF:FF:FF Broadcast address 105 0:9:6B:2:E9:3D IBM Corporation 91 0:4:76:45:61:39 3COM CORPORATION 89 0:0:E2:94:B0:2A Ando Electric Corporation 80 0:6:5B:E6:F8:43 Dell Computer Corp 39 0:C:29:14:1E:63 VMWare, Inc 37 0:0:E2:92:EE:F Ando Electric Corporation 36 0:1:2:79:91:ED 3COM CORPORATION 34 0:50:56:40:0:64 VMWare, Inc 24 1:0:5E:37:96:D0 DoD Internet Multicast (RFC-1112) 24 0:1:3:88:29:92 3COM CORPORATION 22 0:D0:59:C6:5E:14 AMBIT MICROSYSTEMS CORP 20 0:6:5B:D8:BF:ED Dell Computer Corp 11 0:E0:B8:3D:20:40 GATEWAY 2000 11 0:8:74:7:31:EE Dell Computer Corp 10 0:C:29:39:6E:67 VMWare, Inc 9 0:2:A5:B6:E2:E3 Compaq Computer Corporation 8 0:E0:98:A1:7F:DA AboCom Systems, Inc 7 0:B:DB:17:F4:C9 Dell ESG PCBA Test 5 8:0:46:79:F7:7C SONY CORPORATION LTD. 4 0:3:FF:DF:95:84 Connectix 3 1:0:5E:7F:FF:FA DoD Internet Multicast (RFC-1112) 3 0:A0:C9:BA:6D:85 Intel Corporation 2 0:B:DB:DF:53:8D Dell ESG PCBA Test 2 0:A:95:7C:24:0 Apple Computer, Inc. 1 0:8:74:5:B7:F8 Dell Computer Corp 1 0:0:39:F2:67:88 Point Multimedia Systems The target was heavily scanned. IP addresses with the destination MAC 0:50:56:40:0:6d cut -d\ -f3,8 tcpdump.eth.5.txt|sed 's/^\(.* .*\..*\..*\..*\)\..*$/\1/' |grep \ '0:50:56:40:0:6d'|uniq|sort|uniq > 0-50-56-40-0-6d.ip.txt cut -d\ -f 2 0-50-56-40-0-6d.ip.txt | sed 's/\://' | sort | uniq | grep -c "" 1542 cut -d\ -f 2 0-50-56-40-0-6d.ip.txt | sed 's/\://' | sort | uniq | cut -d\. -f1,2,3 | sort -n | uniq -c |sort -nr 256 172.20.11 255 172.22.201 255 172.20.201 255 172.20.12 254 192.168.22 254 192.168.17 1 198.41.0 1 198.123.30 1 172.27.1 1 172.20.102 1 172.11.11 1 172.10.11 1 149.134.52 1 149.134.30 1 134.248.127 1 127.0.0 1 12.162.170 1 10.3.200 1 102.168.17 The alerts from the target MAC were all error responses to bad traffic, except for "Short UDP packet" which was caused by snaplen 96 (the capture length) grep -B2 " 0:50:56:40:0:6D -> " alert | grep "\[\*\*\] " | sort|uniq -c | sort -nr 1326 [**] [1:402:4] ICMP Destination Unreachable (Port Unreachable) [**] 690 [**] [1:399:4] ICMP Destination Unreachable (Host Unreachable) [**] 418 [**] [1:401:4] ICMP Destination Unreachable (Network Unreachable) [**] 34 [**] [1:408:4] ICMP Echo Reply [**] 33 [**] [116:97:1] (snort_decoder): Short UDP packet, length field > payload length 18 [**] [1:449:4] ICMP Time-To-Live Exceeded in Transit [**] 10 [**] [1:491:6] INFO FTP Bad login [**] 7 [**] [1:524:6] BAD-TRAFFIC tcp port 0 traffic [**] 4 [**] [1:451:4] ICMP Timestamp Reply [**] 4 [**] [1:382:4] ICMP PING Windows [**] 2 [**] [1:718:6] TELNET login incorrect [**] 1 [**] [1:404:4] ICMP Destination Unreachable (Protocol Unreachable) [**] The "tcp port 0 traffic" alerts were resets in response to syn packets. egrep " 172\.20\.11\.2\.0 > 10\.10\.10\.141\.|10\.10\.10\.141\..* > 172.20.11.2.0:" tcpdump.eth.5.txt | tail -n 2 14:09:22.899069 0:d0:59:c6:5e:14 0:50:56:40:0:6d 0800 60: 10.10.10.141.62917 > 172.20.11.2.0: S 3868:3868(0) win 512 14:09:22.939627 0:50:56:40:0:6d 0:d0:59:c6:5e:14 0800 60: 172.20.11.2.0 > 10.10.10.141.62917: R 0:0(0) ack 3869 win 0 (DF) A lot of malicious traffic was sent to the target MAC command: grep -B2 " -> 0:50:56:40:0:6D " alert | grep "\[\*\*\] " | sort | uniq -c | sort -nr 18176 [**] [111:9:1] (spp_stream4) STEALTH ACTIVITY (NULL scan) detection [**] 5041 [**] [1:465:1] ICMP ISS Pinger [**] 2144 [**] [1:620:3] SCAN Proxy (8080) attempt [**] 1020 [**] [1:474:1] ICMP superscan echo [**] 545 [**] [116:97:1] (snort_decoder): Short UDP packet, length field > payload length 394 [**] [1:615:4] SCAN SOCKS Proxy attempt [**] 287 [**] [1:1417:2] SNMP request udp [**] 146 [**] [1:1420:2] SNMP trap tcp [**] 140 [**] [1:1418:2] SNMP request tcp [**] 116 [**] [1:1421:2] SNMP AgentX/tcp request [**] 105 [**] [1:618:4] SCAN Squid Proxy attempt [**] 70 [**] [1:1443:2] TFTP GET passwd [**] 60 [**] [1:1411:3] SNMP public access udp [**] 59 [**] [1:467:1] ICMP Nemesis v1.1 Echo [**] 46 [**] [1:1419:2] SNMP trap udp [**] 29 [**] [1:553:4] POLICY FTP anonymous login attempt [**] 27 [**] [1:628:2] SCAN nmap TCP [**] 24 [**] [1:365:5] ICMP PING (Undefined Code!) [**] 24 [**] [111:10:1] (spp_stream4) STEALTH ACTIVITY (XMAS scan) detection [**] 21 [**] [111:12:1] (spp_stream4) NMAP FINGERPRINT (stateful) detection [**] 14 [**] [1:361:7] FTP site exec [**] 12 [**] [1:2049:1] MS-SQL ping attempt [**] 10 [**] [1:1413:2] SNMP private access udp [**] 8 [**] [1:1777:2] FTP EXPLOIT STAT * dos attempt [**] 7 [**] [1:524:6] BAD-TRAFFIC tcp port 0 traffic [**] 7 [**] [1:408:4] ICMP Echo Reply [**] 6 [**] [1:566:3] POLICY PCAnywhere server response [**] 6 [**] [1:528:4] BAD-TRAFFIC loopback traffic [**] 6 [**] [1:1957:3] RPC sadmind UDP PING [**] 6 [**] [1:1893:1] SNMP missing community string attempt [**] 6 [**] [1:1867:1] MISC xdmcp info query [**] 6 [**] [1:1504:5] MISC AFS access [**] 6 [**] [105:1:1] spp_bo: Back Orifice Traffic detected (key: 31337) [**] 4 [**] [1:453:4] ICMP Timestamp Request [**] 4 [**] [1:388:4] ICMP Address Mask Request [**] 4 [**] [1:336:5] FTP CWD ~root attempt [**] 4 [**] [1:237:1] DDOS Trin00:MastertoDaemon(defaultpassdetected!) [**] 4 [**] [1:1992:1] FTP LIST directory traversal attempt [**] 4 [**] [1:1449:3] POLICY FTP anonymous (ftp) login attempt [**] 3 [**] [1:604:5] RSERVICES rsh froot [**] 3 [**] [1:501:2] MISC source route lssre [**] 3 [**] [1:1432:4] P2P GNUTella GET [**] 2 [**] [1:659:4] SMTP expn decode [**] 2 [**] [1:356:5] FTP passwd retrieval attempt [**] 2 [**] [1:332:5] FINGER 0 query [**] 2 [**] [1:330:6] FINGER redirection attempt [**] 2 [**] [1:327:5] FINGER remote command pipe execution attempt [**] 2 [**] [1:326:5] FINGER remote command ; execution attempt [**] 2 [**] [1:323:4] FINGER root query [**] 2 [**] [1:1728:2] FTP CWD ~<CR><NEWLINE> attempt [**] 1 [**] [1:489:5] INFO FTP No Password [**] 1 [**] [1:335:4] FTP .rhosts [**] 1 [**] [1:255:8] DNS zone transfer TCP [**] 1 [**] [1:245:1] DDOS mstream handler ping to agent [**] 1 [**] [1:239:1] DDOS shaft handler to agent [**] 1 [**] [1:236:3] DDOS Stacheldraht client check gag [**] 1 [**] [1:1928:3] FTP shadow retrieval attempt [**] 1 [**] [1:1919:3] FTP CWD overflow attempt [**] 1 [**] [1:1444:2] TFTP Get [**] 1 [**] [111:13:1] (spp_stream4) STEALTH ACTIVITY (SYN FIN scan) detection [**] There are a few alerts that do not involve the target MAC, but they can be ignored. It is possible that the other systems may have been communicating with each other. grep ':..\?:..\?:..\?:..\?:..\?\|^\[\*\*\] ' alert|grep -vB1 '^\[\| 0:50:56:40:0:6D'|grep '^\['|sort|uniq -c 603 [**] [116:97:1] (snort_decoder): Short UDP packet, length field > payload length [ 4 [**] [1:365:5] ICMP PING (Undefined Code!) [**] 4 [**] [1:408:4] ICMP Echo Reply [**] One issue with spoofing is if the target could be used as a dummy for an idlescan[10] The table below indicates that the target may maintain a different IP sequence for each foreign IP and therefore could not be used as a dummy for a spoof scan. However a much more through analysis is required for a real answer.. The target may maintain a different IP sequence for each foreign IP. egrep " 192\.168\.17\.68\..* > " tcpdump.eth.5.txt| grep " ack "| grep -v 'ack [210] win '| cut -d\ -f6-12| tail -n 32|head -n 12 192.168.17.68.80 > 10.10.10.165.4729: R 0:0(0) ack 1180636066 192.168.17.68.80 > 10.10.10.165.4732: R 0:0(0) ack 1181123598 192.168.17.68.80 > 10.10.10.234.1069: R 0:0(0) ack 1730051307 192.168.17.68.80 > 10.10.10.165.4735: R 0:0(0) ack 1181576792 192.168.17.68.80 > 10.10.10.165.4737: R 0:0(0) ack 1182006667 192.168.17.68.80 > 10.10.10.165.4739: R 0:0(0) ack 1182438820 192.168.17.68.80 > 10.10.10.165.4742: R 0:0(0) ack 1182878632 192.168.17.68.80 > 10.10.10.165.4744: R 0:0(0) ack 1183275579 192.168.17.68.80 > 10.10.10.165.4750: R 0:0(0) ack 1184026471 192.168.17.68.80 > 10.10.10.234.1070: R 0:0(0) ack 1732013017 192.168.17.68.80 > 10.10.10.165.4753: R 0:0(0) ack 1184605667 192.168.17.68.80 > 10.10.10.165.4755: R 0:0(0) ack 1185072914 10.10.10.113 was the source for 18202 alerts, Mainly a NULL Scan.of the target. grep -B3 '^10.10.10.113.* -' logs/alert|grep -iB2 '0:50:56:40:0:6d'|grep '\[\*\*\]'|sort|uniq -c 4 [**] [111:10:1] (spp_stream4) STEALTH ACTIVITY (XMAS scan) detection [**] 8 [**] [111:12:1] (spp_stream4) NMAP FINGERPRINT (stateful) detection [**] 18162 [**] [111:9:1] (spp_stream4) STEALTH ACTIVITY (NULL scan) detection [**] 8 [**] [116:97:1] (snort_decoder): Short UDP packet, length field > payload length [**] 4 [**] [1:628:2] SCAN nmap TCP [**] Destination IP's grep '0:50:56:40:0:6d' tcpdump.eth.5.txt|grep "10\.10\.10\.113\..* > "|cut -d\> -f2|cut -d\. -f1,2,3,4|sort|uniq -c|sort -nr Count IP 6626 192.168.17.129 6321 192.168.17.68 5291 192.168.17.135 The number of distinct destination ports scanned for each IP grep '0:50:56:40:0:6d' tcpdump.eth.5.txt|grep "10\.10\.10\.113\..* > "|cut -d\> -f2|cut -d\. -f1,2,3,4,5|sort|uniq|cut -d\. -f1,2,3,4|uniq -c Count IP 1670 192.168.17.129 1657 192.168.17.135 1691 192.168.17.68 The Alert Breakdown grep -B3 '^10.10.10.113.* -> 192.168.17.129' logs/alert|grep -iB2 '0:50:56:40:0:6d'|grep '\[\*\*\]'|sort|uniq -c 1 [**] [111:10:1] (spp_stream4) STEALTH ACTIVITY (XMAS scan) detection [**] 2 [**] [111:12:1] (spp_stream4) NMAP FINGERPRINT (stateful) detection [**] 6607 [**] [111:9:1] (spp_stream4) STEALTH ACTIVITY (NULL scan) detection [**] 2 [**] [116:97:1] (snort_decoder): Short UDP packet, length field > payload length [**] 1 [**] [1:628:2] SCAN nmap TCP [**] grep -B3 '^10.10.10.113.* -> 192.168.17.68' logs/alert|grep -iB2 '0:50:56:40:0:6d'|grep '\[\*\*\]'|sort|uniq -c 3 [**] [111:10:1] (spp_stream4) STEALTH ACTIVITY (XMAS scan) detection [**] 6 [**] [111:12:1] (spp_stream4) NMAP FINGERPRINT (stateful) detection [**] 6264 [**] [111:9:1] (spp_stream4) STEALTH ACTIVITY (NULL scan) detection [**] 6 [**] [116:97:1] (snort_decoder): Short UDP packet, length field > payload length [**] 3 [**] [1:628:2] SCAN nmap TCP [**] grep -B3 '^10.10.10.113.* -> 192.168.17.135' logs/alert|grep -iB2 '0:50:56:40:0:6d'|grep '\[\*\*\]'|sort|uniq -c 5291 [**] [111:9:1] (spp_stream4) STEALTH ACTIVITY (NULL scan) detection [**] Syn's were sent to ports 1 and 20 of 192.168.17.129 and 192.168.17.68. No Syn's were sent to 192.168.17.135 command grep '0:50:56:40:0:6d' tcpdump.eth.5.txt|grep " S "|grep "10\.10\.10\.113\..* > "|cut -d\ -f8,9|sort|uniq -c Count IP 6 192.168.17.129.1: S 1 192.168.17.129.20: S 18 192.168.17.68.1: S 3 192.168.17.68.20: S It looks like every open port will respond with a reset to a null scan. grep '0:50:56:40:0:6d .* > 10\.10\.10\.113\.' tcpdump.eth.5.txt|cut -d\ -f6,9|egrep -v "68\.20 |129\.20 "|sort|uniq -c|sort -nr 1 192.168.17.68.80 R 1 192.168.17.68.53 R 1 192.168.17.68.443 R 1 192.168.17.68.25 R 1 192.168.17.68.23 R 1 192.168.17.68.22 R 1 192.168.17.68.21 R 1 192.168.17.135.53 R 1 192.168.17.135.22 R 1 192.168.17.135.20 R 1 192.168.17.129.80 R 1 192.168.17.129.53 R 1 192.168.17.129.443 R 1 192.168.17.129.25 R 1 192.168.17.129.23 R 1 192.168.17.129.21 R 192.168.17.135 answering to a null scan from10.10.10.113. grep "10\.10\.10\.113\." tcpdump.eth.5.txt|grep "192\.168\.17\.135\."|egrep "135\.53[ :]|135\.20[ :]|135\.22[ :]"|sort|uniq 14:14:55.665504 0:a:95:7c:24:0 0:50:56:40:0:6d 0800 60: 10.10.10.113.59194 > 192.168.17.135.22: . win 4096 14:14:55.671788 0:50:56:40:0:6d 0:a:95:7c:24:0 0800 60: 192.168.17.135.22 > 10.10.10.113.59194: R 0:0(0) ack 0 win 0 (DF) 14:15:51.827656 0:a:95:7c:24:0 0:50:56:40:0:6d 0800 60: 10.10.10.113.59194 > 192.168.17.135.53: . win 2048 14:15:51.854980 0:50:56:40:0:6d 0:a:95:7c:24:0 0800 60: 192.168.17.135.53 > 10.10.10.113.59194: R 0:0(0) ack 0 win 0 (DF) 14:16:10.128439 0:a:95:7c:24:0 0:50:56:40:0:6d 0800 60: 10.10.10.113.59194 > 192.168.17.135.20: . win 1024 14:16:10.132772 0:50:56:40:0:6d 0:a:95:7c:24:0 0800 60: 192.168.17.135.20 > 10.10.10.113.59194: R 0:0(0) ack 0 win 0 (DF) Three of the hosted VM's . could be Linux systems. The TTL discrepancy could be caused by a separate VM on the target that that is configured as a firewall. cmd grep -A2 ' 0:50:56:40:0:6D -> ' logs/alert|grep ' TCP .* IpLen:'|cut -d\ -f1,4,5|sed 's/:.. T/ T/' |sort|uniq -c|sort -k2 7 172.20.11.2:0 TCP TTL:62 4 172.20.201.135 TCP TTL:62 8 172.20.201.198 TCP TTL:62 Attack Mechanism Here is a brief description of the idle scan that was invented by Antirez[8] The goal of the idle scan is that no traffic travels between the attacker and the target. The attacker sends a non spoofed packet to a third system (the dummy) that uses sequential IP ID's. The attacker then reads the IP ID from the reply. Next the attacker sends a spoofed packet were the source field contains the IP of the real target. The dummy system responds with an error message to the target. If the target has the same port open it should respond otherwise is should not respond. A response would increment the IP ID counter on the dummy. Then the attacker sends a non-spoofed packet to the dummy machine and reads the IP ID. From the difference of the IP ID's he can tell of the target responded or not. If the third system is not busy then many ports can be queried before the attacker has to use a non-spoofed packet. Fig 2.1.2 indicates that the target may maintain an independent IP sequence for each foreign IP. Therefore it could not be used as a dummy for an idle scan. However a more through analysis is required for a definitive answer A null scan is a TCP packet with no flags. When it is sent to an Open port a reset may be sent back but would not be entered into the hosts logs. A closed port should not respond. This way an attacker can scan for open ports without detection. Some older firewalls would not drop a null TCP packet. Traffic to and from port 0 is not allowed in the RFC's. Therefore different OS's respond differently. It can be used for OS typing. Correlations: Several other people have posted detects from the same source file to intrusions@xxxxxxxxxxxxxx They all came to different results because they did not consider a pen test. I could not find a CVE entry for a Null Scan [11] Evidence of active targeting: The attacks were mainly to three specific IP's Severity: severity = (criticality + lethality)-(system countermeasures + network countermeasures) = (5 + 5) - (3 + 0) = 7 criticality: 5 A Vmware server may hold several virtual servers. If compromised then all may be compromised lethality 5 If an attack of this size happened on a production LAN, I would turn everything off. Normal Business would stop and the business continuity plan would be invoked. system countermeasures 3 The target was not broken into, but it did responded to the null scan and port 0 traffic. network countermeasures 0 Nothing was blocked from the VMware server. Defensive recommendation: First, none of this traffic should be allowed to get to the Internet or a production LAN. There are several critical unknowns with this analysis. VMWare hosts other OS's but they are not known. There may have been a separate firewall on the target and/or the hosted systems had some sort of firewall. The target configuration was unknown. The target may be dual homed. The target seemed to do OK but there is room for improvement. Drop all traffic to port 0. Traffic to and from port 0 can be used for OS typing. This would be useful to an attacker and probably not much use to IT. Drop all null TCP packets. The target answered to a null scan. There is no legal reason for this traffic. The error messages from the target would be useful to an attacker or a legitimate SA. It would increase security to disable the error responses, and the LAN would still work, but it would be much harder to diagnose and repair problems. One option would be to set up one of the virtual machines as a firewall/router and log the errors. Then the SA would be able to see the messages but an attacker would not. Multiple choice test question: Which IP range includes IP's that are legal on the Internet. A) 10.0.0.0 - 10.255.255.255. B) 192.168.0.0 - 192.168.255.255. C) 172.16.0.0 - 172.31.255.255 . D) 224.0.0.0 - 239.255.255.255 E) none Answer E: All subnets are reserved by ICANN. See http://whois.arin.net 2.1.11 References [1] Spirent Communications. Security & Web Infrastucture. URL: http://www.spirentcom.com/analysis/index.cfm?WS=65 (2 May 2004) [2] Ando Electric Co., Ltd. URL: http://www.ando.com (20 March 2004). [3] Cahners Publishing Company. Electronic News: CADENCE TO ACQUIRE AMBIT?(Cadence Design Systems considers acquisition of Ambit Design Systems) (Company Business and Marketing). 24 August 1998. URL: http://www.findarticles.com/cf_dls/m0EKF/n2233_v44/21054771/p1/article.jhtml (20 March 2004). [4] AboCom Systems , Inc. URL: http://www.abocom.com/ (20 March 2004). [5] VMware, Inc. URL: http://www.vmware.com/support/ (20 March 2004). [6] Heim, Kristi. Mercury News Microsoft buys assets of Connectix. 20 Feb. 2003. URL: http://www.siliconvalley.com/mld/siliconvalley/5222197.htm (20 March 2004). [7] Sony Corporation of America. URL: http://sony.com/ (20 March 2004). [8] Cavebear. Ethernet Codes: Multicast (including Broadcast) Addresses. 09 March 1999. URL: http://www.cavebear.com/CaveBear/Ethernet/multicast.html (20 March 2004). [9] IEEE OUI and Company_id Assignments URL: http://standards.ieee.org/regauth/oui/index.shtml (20 March 2004). [10] Vaskovich, Fyodor. Idle Scanning and related IPID games.13 November 2003. URL: http://www.insecure.org/nmap/idlescan.html (20 March 2004). [11] URL: http://cve.mitre.org (20 March 2004). _______________________________________________ Intrusions mailing list Intrusions@xxxxxxxxxxxxxx http://www.dshield.org/mailman/listinfo/intrusions |
|
| <Prev in Thread] | Current Thread | [Next in Thread> |
|---|---|---|
| Previous by Date: | [LOGS] Summary of large-scale portscanning detects: 00005, Ken . Connelly |
|---|---|
| Next by Date: | [LOGS] Summary of large-scale portscanning detects: 00005, Ken . Connelly |
| Previous by Thread: | LOGS: GIAC GCIA Version 3.4 Practical Detect- David Chance (2nd Attempt)i: 00005, skycopp-mail8531 |
| Next by Thread: | Excessively large URI attacks: 00005, Barry Fitzgerald |
| Indexes: | [Date] [Thread] [Top] [All Lists] |
| News | FAQ | advertise |