Sponsored by: █ Sparkhost - Hosting Without Compromises! █ Hybrid Performance Web Hosting █ Spark Host Stream Hosting █ Hybrid IRC & IRCd Server Shell Accounts
Pwdump Series
#1
Posted 17 August 2010 - 11:58 PM
What would be useful would be tools and techniques for a standard domain user to be able to get hashes from remote servers without administrative credentials? Do such tools/techniquse exisit (considering the target server is Windows 2003 and is fully patched), if so please point me in the direction, and more importantly and mitigating controls to protect such attacks. Physical access to the Server is one option I know with boot CD's and distros - but I'm after something where physical access is not possible...
#2
Posted 18 August 2010 - 02:46 AM
or maybe for using the admin account for keeping control rather than creating a new account etc.
Getting hashes of remote servers would be considered a vulnerability. So, No! You won't get those.
You could try pass-the-hash technique. Tools like gsecdump, incognito should help
I found a link for you as well
http://carnal0wnage....ito-part-2.html
#3
Posted 20 August 2010 - 12:57 AM
#4
Posted 06 October 2010 - 06:31 PM
I guess you answered the first question yourself. You could also use those passwords on the network to see if it is being repeated
or maybe for using the admin account for keeping control rather than creating a new account etc.
Getting hashes of remote servers would be considered a vulnerability. So, No! You won't get those.
You could try pass-the-hash technique. Tools like gsecdump, incognito should help
I found a link for you as well
http://carnal0wnage....ito-part-2.html
I haven't tried this tool as of yet, but do you need to have the AV turned off or it won't matter.
#5
Posted 06 October 2010 - 11:55 PM
#6
Posted 07 October 2010 - 09:30 AM
Read the rules before you post
#7
Posted 07 October 2010 - 11:22 AM
Do such tools/techniquse exisit (considering the target server is Windows 2003 and is fully patched), if so please point me in the direction, and more importantly and mitigating controls to protect such attacks. Physical access to the Server is one option I know with boot CD's and distros - but I'm after something where physical access is not possible...
One mitigation control is to prevent Windows from storing LM hashes. When a password is longer than 14 characters, no LM hash is stored.
And there are settings to prevent Windows from storing a LM hash for passwords shorter than 15 characters, take a look at the Microsoft KB article:
http://support.microsoft.com/kb/299656
#8
Posted 07 October 2010 - 11:27 PM
Do such tools/techniquse exisit (considering the target server is Windows 2003 and is fully patched), if so please point me in the direction, and more importantly and mitigating controls to protect such attacks. Physical access to the Server is one option I know with boot CD's and distros - but I'm after something where physical access is not possible...
One mitigation control is to prevent Windows from storing LM hashes. When a password is longer than 14 characters, no LM hash is stored.
And there are settings to prevent Windows from storing a LM hash for passwords shorter than 15 characters, take a look at the Microsoft KB article:
http://support.microsoft.com/kb/299656
By default that is not necessary in Windows Vista or 7, only in Windows XP or past versions.
#9
Posted 07 October 2010 - 11:36 PM
By default that is not necessary in Windows Vista or 7, only in Windows XP or past versions.
Correct, but the OP was asking about Windows Server 2003, and this KB-article applies to Windows Server 2003.
#10
Posted 08 October 2010 - 02:36 AM
By default that is not necessary in Windows Vista or 7, only in Windows XP or past versions.
Correct, but the OP was asking about Windows Server 2003, and this KB-article applies to Windows Server 2003.
Got a bit carried away, when I saw the explanation on lm and ntlm hashes and didn't notice the Windows Server 2003 part. My apologies there.
#11
Posted 16 November 2010 - 11:07 PM
I use PWDump/FGDump on almost every penetration test I do.
It is very frequent that I gain access to a single server. In an environment with 500 machines, it's actually ordinary.
However, escalating that to the rest of the environment is the issue.
With FGDump, I can both crack the local admin password, which is often repeated on other machines, AND i can capture cached domain credentials.
A typical scenario is one I had last week. I cracked a single SQL 2000 box, and used PWDump to get the admin password hashes and cracked them via rainbowcrack. Then I ran a tool that queries other machines to see if they use the same password. I found 114 machines that did.
At this point, I have a perl script I wrote that runs a batch process on those machines, both dumping the local SAM passwords, as well as the domain cached credentials. I gathered 210 cached domain user accounts, and 10 additional common local user name/password combinations.
From these local users with LM hahses, I was able to penetrate an additional machine, revealing a second (newer) local administrator password, yielding another 140 systems. Now i have almost 400 cached domain credentials in a salted mscache format. Running this through a dictionary cracker, I gain access to 85 domain accounts, including one domain administrator. From there, I have complete access to the entire infrastructure.
As you can see, the SQL exploit, mixed with pwdump, combined with rainbowcrack and john the ripper, turned access to a single machine into completed administrative control of the domain in about 6 hours work and 40 hours of computational time.
Win!
#12
Posted 21 December 2010 - 07:49 AM
Do such tools/techniquse exisit (considering the target server is Windows 2003 and is fully patched), if so please point me in the direction, and more importantly and mitigating controls to protect such attacks. Physical access to the Server is one option I know with boot CD's and distros - but I'm after something where physical access is not possible...
One mitigation control is to prevent Windows from storing LM hashes. When a password is longer than 14 characters, no LM hash is stored.
And there are settings to prevent Windows from storing a LM hash for passwords shorter than 15 characters, take a look at the Microsoft KB article:
http://support.microsoft.com/kb/299656
By default that is not necessary in Windows Vista or 7, only in Windows XP or past versions.
I'm also working with FGdump and PWDump, but I never used it on a Windows 7 box.
Is windows 7 using the same technique as Windows XP? And if not, wich kind of hashing is it using and how to extract them on a Windows 7/Vista box..
#13
Posted 21 December 2010 - 10:54 AM
http://www.openwall....xp-2003-vista-7
It's just that the hashes are of NTLM format and not LM.
0 user(s) are reading this topic
0 members, 0 guests, 0 anonymous users












