Bugtraq mailing list archives
Buglet in Bind 4.9.5
From: alan () MANAWATU GEN NZ (Alan Brown)
Date: Thu, 10 Apr 1997 17:36:20 +1200
I just got tripped up by this. I wonder how it's managed to be undetected for so long.... If an A record is assigned with leading zeros to pad out the number and make it slightly easier to throw through sort, the assigned number goes wonky. Eg. fubar IN A 202.50.64.125 = 202.50.64.125 fubar IN A 202.050.64.125 = 202.40.52.125 fubar IN A 1.2.3.4 = 1.2.3.4 fubar IN A 001.002.003.004 = 1.2.3.4 As can be seen, it only happens with one padding 0. If there are 2 it's fine and (I assume) that's why it's not been notified up till now. Security implications? I use static assignments for my dialin users and they were setting off stacks of alarms as they accessed the pop port until I realised what'd happend and back out the changes. The other problem was that IPs which had been locked from getting offsite may have been able to get out, but also they were so whacked out that nothing was coming back. I assume that someone else in the 202.40.5? block was seeing sporadic bogus replies and may have had alarms tripped too. AB
Current thread:
- qualcomm POP server David Sacerdote (Apr 09)
- Buglet in Bind 4.9.5 Alan Brown (Apr 09)
- Buglet in Bind 4.9.5. [SUMMARY] Alan Brown (Apr 10)
- CIAC Bulletin H-45: Windows NT SAM permission Vulnerability Aleph One (Apr 10)
- Norton Utilities 2.0 Vulnerability Aleph One (Apr 10)
- L0pht Advisory: release of L0phtCrack for NT Aleph One (Apr 11)
- New source address for Sun Security Bulletins Aleph One (Apr 11)
- [LINUX] IP_MASQ / Ethernet Passing Traffic After Halt Sean B. Hamor (Apr 11)