While researching an IRC zombie infection for a third party, I came across a variant of sdbot which is hard-coded to send TCP packets with a window size of 55808 in its spoofed-synflooding function. Could this be the "next-gen" trojan that Lancope found? If so, it redefines the term next-gen, because sdbot is pretty much old-school. The particular variant I found is _not_ the well known "sdbot SYN edition" by Tesla. The packet construction subroutine is entirely different. Based on what I have seen here and in the firewall logs from as far back as January, I feel that there is a snippet of C being re-used in the underground and it uses a default window size of 55808. I've seen it used in broadscanning, and now as a synflooder. Here is an example command used in the IRC control channel to start a synflood with this version of sdbot. 192.168.1.21 is the address to be spoofed while attacking 192.168.1.1: $syn 192.168.1.1 6000 20 192.168.1.21 6666 Here is a capture of some of the resulting packets: 07:26:51.048897 192.168.1.21.6666 > 192.168.1.1.6000: S 693933104:693933104(0) win 55808 0x0000 4500 0028 0a34 0000 8006 ad35 c0a8 0115 E..(.4.....5.... 0x0010 c0a8 0101 1a0a 1770 295c 9430 0000 0000 .......p)\.0.... 0x0020 5002 da00 6374 0000 0000 0000 0000 P...ct........ 07:26:51.049000 192.168.1.21.6666 > 192.168.1.1.6000: S 3950185482:3950185482(0) win 55808 0x0000 4500 0028 0a35 0000 8006 ad34 c0a8 0115 E..(.5.....4.... 0x0010 c0a8 0101 1a0a 1770 eb73 0c0a 0000 0000 .......p.s...... 0x0020 5002 da00 2983 0000 0000 0000 0000 P...)......... 07:26:51.049096 192.168.1.21.6666 > 192.168.1.1.6000: S 2692113931:2692113931(0) win 55808 0x0000 4500 0028 0a36 0000 8006 ad33 c0a8 0115 E..(.6.....3.... 0x0010 c0a8 0101 1a0a 1770 a076 660b 0000 0000 .......p.vf..... 0x0020 5002 da00 1a7f 0000 0000 0000 0000 P............. I have passed the binaries I have found along to the AV community, so anti-virus signatures at least for the variants I have found should be forthcoming. Of course, this still doesn't explain the weird source and destination IP addresses and ports we are seeing since last month, but based on this I seriously doubt it is a covert channel. Maybe someone is just testing a new implementation of the synscanning code in a distributed manner, and has some bugs to work out. -Joe -- Joe Stewart, GCIH Senior Intrusion Analyst LURHQ Corporation http://www.lurhq.com/ ---------------------------------------------------------------------------- Attend the Black Hat Briefings & Training, July 28 - 31 in Las Vegas, the world's premier technical IT security event! 10 tracks, 15 training sessions, 1,800 delegates from 30 nations including all of the top experts, from CSO's to "underground" security specialists. See for yourself what the buzz is about! Early-bird registration ends July 3. This event will sell out. www.blackhat.com ----------------------------------------------------------------------------
This archive was generated by hypermail 2b30 : Wed Jun 18 2003 - 08:39:03 PDT