Open GoogleCodeExporter opened 9 years ago
Hey..
a few questions/requests:
1) Can you send me a wireshark dump of that attack? Try to constrain just the
hosts involved for better understanding
2) What payload.exe is?
From the output you sent me, looks like the authentication succeeded (the most
important part) but there was a problem trying to request the shares on the
other side.. which is kind of weird.. It says tho, that the service ended up
installed .. but I'm assuming nothing happened on your side. The wireshark dump
will help me a lot.
thanks for reporting this!
beto
Original comment by bet...@gmail.com
on 10 Dec 2013 at 6:59
Hi
in attach the dump from wireshark, showing the browsing to \\ip_attack\admin$.
please keep in mind that the share is on the DC(246.207)the machines have no
antivirus and that i did not change any GPO.
the payload.exe its a meterpreter_reverse that i have also tested on the DC and
client by copy it manually to them and it works fine.
Thks for the help.
F.
Original comment by clip0c...@gmail.com
on 11 Dec 2013 at 12:33
Attachments:
Hey Mate:
Thanks for the dump. Just to clarify whether I'm right:
1) 192.168.246.205 is the victim machine, where you're browsing
\\ip_attack\\admin$ from
2) 192.168.246.211 is the attacking machine, where you will forward
192.168.246.205 credentials to the target host
1) 192.168.246.207 is the target host you want to relay the credentials to
If i'm right with that.. take a look at packet 96.
The target machine is accepting the authentication BUT, it is signing SMB
Packets. Then afterwards attacking machine wants to connect to
\\192.168.246.207\ipc$ and it gets access denied ('cause the was no signature
in the original packet).
Now.. it could be two things:
1) The server is enforcing signing (can you check the local security policies
of the target machine?, LOCAL POLICIES->SECURITY OPTIONS->Microsoft network
server: Digitally sign communications* and tell me the values you have?)
2) The SMB Relay module is not clearing the bit for signing, hence shooting
himself in the foot. I gotta take a look at the code.. then I'll let you know.
thanks for the report!
beto
Original comment by bet...@gmail.com
on 11 Dec 2013 at 4:47
Also check, at the victim machine, the policies for LOCAL POLICIES->SECURITY
OPTIONS->Microsoft network client: Digitally sign communications*
Original comment by bet...@gmail.com
on 11 Dec 2013 at 5:04
Take a look at packet 87, 192.168.246.207 is saying signatures are required :(
You can find that data, inside the packet, Negotiate Protocol Response ->
Security Mode.
Most probably the domain is enforcing signing of smb packets. If so, this
attack cannot be performed.
Please let me know if this is right.. otherwise we'll keep researching.
cheers
beto
Original comment by bet...@gmail.com
on 11 Dec 2013 at 5:10
Hi beto,
you are 100% correct in all.
DC and Domain GPO forcing digital signing of smb packets. But client only siggn
if server enforce!
So no work arrow this?
Thks.
F.
Original comment by clip0c...@gmail.com
on 11 Dec 2013 at 6:19
Yep.. so the server will reject packets if they are not signed..
Read the smbrelayx.py comments at the beginning:
# The only way to stop this attack is to enforce on the server SPN checks and
or signing.
So.. nope.. with those options on you cannot relay connections.
Just in case, could you disable the server's configuration Microsoft network
server: Digitally sign communications(always) and perform the attack against to
see if it works?
thanks mate.
beto
Original comment by bet...@gmail.com
on 11 Dec 2013 at 7:04
Yeap.. disabeling on the domain controler GPO the Microsoft network server:
Digitally sign communications(always) allows the attack to work in perfection.
thks beto.
Original comment by clip0c...@gmail.com
on 12 Dec 2013 at 12:50
Thanks for checking it out mate.
I'm gonna add a warning message whenever the target server has signing required
so the user knows before hand.
cheers
beto
PS: Stay tuned, I have some ideas that might workaround this limitation.. but I
still gotta do some research ;)
Original comment by bet...@gmail.com
on 12 Dec 2013 at 1:24
Original issue reported on code.google.com by
clip0c...@gmail.com
on 10 Dec 2013 at 4:57