aeturnus
-
Posts
65 -
Joined
-
Last visited
Posts posted by aeturnus
-
-
Well this may sound like a noob question, but I'm currently running Linux, so I am not able to try this out myself. I have looked a little at the SysInternal tool "PsTools" for remotely running commands, on different computers. I can easily so how I am suppose to use it, but does anyone know what it requires on the other side, I guess you would need some kind of access to the remote user, to avoid you from just running whatever en an entire network, or is it simply just the script kitties wet dream?
Anybody had any experience with it?
I think this has been mentioned, but you're looking for "psexec". To run it you require Administrator privileges on the remote computer and access to the remote computers ADMIN$ share.
-
Blacklisting is the correct way to stop unwanted behavior. Whitelisting is when you assume the users are evil little buggers from the get go. If you blacklist the major trackers plus keep an eye on new ones that are been used you significantly reduce the chance of his torrent client working.
If his client supports DHT that would be a bit more difficult to stop entirely but you can certainly cripple it with even a retarded firewall.
Did that even make sense in your head?
You want him to try to document all the trackers this guy's going to be using and hope he's not using DHT? Who doesn't use Azureus with encrypted transport? (DHT + no packet signature by default).
Blacklisting is not the answer for this. The elegant solution allows him to have the greatest control over the other guy's network speeds, not hoping that the other guy isn't as good as the vast majority of users on the Internet.
-
If you're going the SSH route, don't forget to tunnel your DNS traffic, otherwise it's all for naught.
-
Hopefully the firewall would also have the feature to spot bittorrent traffic by packet signature. Even with out this feature, all you have to do is use the firewall to build a list of trackers IP addresses and black list them.
A hardware firewall device is the correct answer in this situation I believe. You might be able to do a ARP poison the switch(s) and then possibly spoof a bunch of reset packets for every request his bitorrent client makes, but you might not either. For example, his computer might be running a firewall such as Zone Alarm (even though I never would use this crapware) which prevents this. If this is the case all bets are off, get a hardware firewall.
Again, where's he going to put the hardware firewall? In between his connection and the router would do no good as it would have no effect on the rogue user. He all ready stated that he has no access to the router (which seems to imply where their connections come from), so it seems unlikely that he'd be able to just wander into the switch cabinet and add a device.
However, if he were to make himself the rogue user's gateway (through whatever means you'd like -- ARP Poisoning was just an example; and what percentage of users use static ARP tables?) then this is a nonissue all together since he would effectively be the administrator of the rogue client's network and would get the same benefits of a hardware firewall.
Further, blacklisting, in general is silly. Read a security book. You want to whitelist. And what packet signature would you expect the firewall to pick up if the Bittorrent client is using encrypted transport? Random bytes to random IPs on random ports is a ridiculous idea to try to blacklist.
-
I enjoyed learning from Bjarne Stroustrup's book The C++ Programming Language. I think the 3rd Edition is out now.
-
Get a firewall, put it in between the 'modem' and the reset of the network, block all ports you don't need out bound, which for web browsing will be 80, 443 and 53.
It doesn't sound like this is possible given his situation. If he could physically add a firewall between the gateway and the rest of the nodes, he could physically unhook the cable.
There's really not enough information here to answer his question completely, but Deauthing probably won't work either as he doesn't really imply there's any authentication needed to use the LAN.
Further, he doesn't allude to it being a corporate situation. So there's probably not a "boss" to report him to, and it sounds like he did report it to who he thought could help.
Listening to the question would probably help before trying to give advice.
If we take his question as completely hypothetical, or in a controlled lab environment for educational purposes, one might suggest looking at routing protocols. Ideally, if you want this "rogue" user to stop using more than his fair share of the bandwidth, you could become his gateway through any number of methods (you're on the same LAN, look up LAN attacks including ARP cache poisoning). From there you should set up your computer to properly act as his gateway and just throttle his connection speeds for the given Bittorent port.
This way no one is DOS'd, and you still get to call home.
-
MIPS assembly is probably the simplest I know. The easiest for me to pick up was probably C#.
-
I was wanting to read up on some certain parts of windows OS's and I was wondering if Microsoft made some type of "complete guide" that explained pretty much everything like the registry, command prompt, file system, interface, etc.
Depending on what exactly you're looking for, you might find the Windows Internals book interesting. It covers a lot of the details of the inner workings of the operating system. It depends on how low-level you want to go though. If you're interested in kernel debugging, the Advanced Windows Debugging book is pretty decent as well.
-
putty still uses cmd though, am I wrong?
I need a way for people with almost NO technolust ;) to be able to use this. I wouldn't mind showing them how to use putty or another simple app to use it but it needs to be done without access to cmd
I'm not really sure what you mean by "without access to cmd".
If you mean, you don't want the end users to enter any commands into a shell, then you can allow users to log in using certificates and have their computer connect on start up. Doing this will allow them to SSH into your box without any user interaction.
If you're coming from a Windows perspective and you don't want them to use the "cmd" prompt, then Putty has a GUI. I believe plink is the CLI version.
-
Do not be so dismissive of my question. Yes, I've already reviewed Nvidia's website and I've read several CS papers.
The information that I'm seeking is how CUDA is used for cracking and if there are any good aps that do this.
Pyrit looks to be interesting. Google is pretty interesting too.
-
Just jump in!
The water isn't as cold as you believe it is..
Don't try Ubuntu or OpenSUSE..
Not gonna help you a bit.
Gonna recommend RHEL and it's variants.
I don't think I understand your aversion to Debian-based GNU/Linux. My CentOS runs Gnome just like my Ubuntu (well, not true, I use Fluxbox, but I'm pretty sure both are Gnome by default). I think yum is a lot slower than apt-get. And don't get me started on rpm-hell.
As for saying Ubuntu isn't where the big boys play, whynot? I chose Ubuntu for learning to write kernel modules because of its easy bundling my kernel as a package to put on other Ubuntu systems.
Difficult does not mean hardcore.
For the topic at hand, clearly I recommend Ubuntu. But I don't recommend dual-booting (I hate having to reboot jus to get to my other OS), just use a virtual machine.
-
This bot is intended to say "Darren is Awesome!" every so often!
CODE:
import sys import socket import string import os #not necassary but later on I am going to use a few features from this HOST='chat1.ustream.tv' #The server we want to connect to PORT=6667 #The connection port which is usually 6667 NICK='darrenbot' #The bot's nickname IDENT='awesomebot' REALNAME='Darrren Is Awesome' OWNER='dylanwinn' #The bot owner's nick CHANNELINIT='#hak5editcam' #The default channel for the bot readbuffer='' #Here we store all the messages from server def syscmd(commandline,channel): cmd=commandline.replace('sys ','') cmd=cmd.rstrip() os.system(cmd+' >temp.txt') a=open('temp.txt') ot=a.read() ot.replace('n','|') a.close() s.send('PRIVMSG '+channel+' :'+ot+'n') return 0 s=socket.socket( ) #Create the socket s.connect((HOST, PORT)) #Connect to server s.send('NICK '+NICK+'n') #Send the nick to server s.send('USER '+IDENT+' '+HOST+' bla :'+REALNAME+'n') #Identify to server s.send('JOIN '+CHANNELINIT+'n') #Join a channel while 1: if(line[0]=='PING'): #If server pings then pong s.send('PONG '+line[1]+'n') time.sleep(60) s.send('Darren is awesome')ERROR:
IDLE 3.0 >>> ================================ RESTART ================================ >>> Traceback (most recent call last): File "C:/Python30/bot", line 28, in <module> s.send('NICK '+NICK+'n') #Send the nick to server TypeError: send() argument 1 must be string or buffer, not str >>>The error doesn't seem to make sense! Isn't a string the same a STR?
I haven't used Py3k either, but Google tells me it's a bug in that version and you should be using a byte array instead of the str type (http://bugs.python.org/issue4275). Initially I was curious if just wrapping the argument in str() would fix it, but I guess not.
Also, where does your program call recv? From this code, it looks like you're never going to PONG and you'll get disconnected. And should those 'n' be '\n' Or is that forum magic that removed them?
-
You should run this command about once a week:
It does all sorts of cleanings and removes a bunch of stuff.
HAHA.
No, seriously do not run that command unless you work for the government, it basically nukes the root directory.
Read this for commands NOT to run:
http://www.junauza.com/2008/11/7-deadly-linux-commands.html
Also, other things to read
http://www.unixguide.net/linux/linuxshortcuts.shtml
Slightly off topic, but modern *nix operating systems don't let you simply "sudo rm -rf /". You have to do something like:
# cd /
# rm -rf .
As for the actual question: The daily work of a system admin changes quite a bit based on what your company does and the size of the company. In general though, you'd probably be expected to be able to do simple maintenance of a Linux server, upgrading the software, and so forth. You probably won't dive too much into software development beyond simple scripts to make your life easier.
-
Not 100% sure if you need a Squid proxy to handle the request on the hosts end though,
But shouldn't be to hard to configure.
Read up on "-D" option in the manual pages for SSH.
No, you don't need to set up Squid. He also won't need to give his root password to anyone.
If you plan to do this from school or some place that you don't have Linux available to you, I'd suggest checking out Putty for your SSH client. It's just a simple exe that you can download and run. You will still need to specify a SOCKS proxy in whatever application you're trying to run though.
full raw sockets in windows xp?
in Applications & Coding
Posted
Sorry, but it's my understanding that to get your required functionality out of XP SP2 or greater, you'd need a driver to hook on the NDIS or TDI level ( I don't remember which for this case ). There's code examples on the web for this if you need it.