Results 1 to 13 of 13

Home and Office Computer Security Basics

  1. #1
    Member Jake's Avatar
    Join Date
    Nov 2008
    Location
    Dorset UK

    Home and Office Computer Security Basics

    Introduction:

    Pretty much everyone understands the basic point of a password; a word known only to yourself, which controls access to your files or forum account or whatever. You probably had a few run-ins with your school's Head of IT/Comp Sci or Information Technology Manager or whatever fancy title they give the most senior -sometimes the only- faculty member who knows how to fix the computers until you learned to pick a password that wasn't guessable, and stop typing it in the instant you saw someone peering over your shoulder. But out here on the grown-up world, you're up against more than just some annoying little boy with nothing better to do than make your life slightly difficult, and password security is a much bigger issue.
    I'm going to start by explaining some of the tricks you'll be coming up against, then I'll tell you how not to get screwed over by them.

    Typical Avenues of Attack:

    Brute Force, or 'Dictionary' Attacks:
    The least sophisticated form of attack, by which a would-be cracker uses a simple program to fill the password box with a string of random letters and numbers until it hits the right combination. More sophisticated examples run through lists of existing words -names, TV shows etc- before resorting to random strings, but it's still a time-consuming process, even more so for servers that allow only a limited number of login attempts before blocking access for a while. Nevertheless, the longer and more complicated your password is, the more you reduce the odds of them getting it right the first time.

    Phishing:
    In its simplest form, an email purporting to be from the forum administrators or your bank etc, claiming they need your username and password to perform some system maintenance task or other. These aren't especially common now, thanks to increased public awareness that under no circumstances will a legitimate organisation ever actually need to ask you what your password is.
    A more subtle variation involves an email purportedly from a site like Amazon with a link to a special offer, which will lead to a fake login page that captures your details. The best defence against these is to always look at the address bar before logging in when you follow such a link; the URL of the real site is the one thing that can't be faked.

    Social Engineering:
    This is a very broad term that covers a lot of things, but essentially it means obtaining illegitimate access to sensitive data by face-to-face persuasion. Basic social-engineering tricks include covertly watching someone type in their password, or asking to use someone else's PC on some pretext and grabbing stuff while their back is turned. A more audacious tactic is to steal or fake up an ID card, don a suitable disguise and simply brazen your way into the building, and then out again whilst carrying the file server! (Before you dismiss this as implausible, ask yourself just how hard you'd look at some guy in overalls carrying something heavy on a parcel trolley if you passed him in the corridor at work.)
    Obtaining someone's username and password by bribery or coercion is also technically social engineering, albeit a variety that you are fairly unlikely to experience in real life.

    Hacking:
    I mention this avenue of attack only for the sake of completeness, as there's not a lot that you the user can do to prevent this. Your password is stored in a database on a server, often in plain text format, and itself protected only by a password and -hopefully- a firewall. If an unauthorised person gains direct access to that server by some means, they can copy the entire database over to their own machine. The person who runs the server is the one who'll get fired and sued for this rather than you, but all the same, prevention is better than cure; any data you have that could be harmful in the wrong hands should stay off browser-based online storage services like Google Documents.

    General Advice:
    Six characters is a pretty reasonable minimum password length for everyday use such as on forums like this one; hijack attempts are liable to be opportunistic and low-intensity, and the potential for damage fairly limited. Provided you don't pick anything stupidly obvious like 'password' or your username, you can probably get away with something fairly simple like your dog's name or your favourite hero unit from DOW or whatever. Something like your Paypal account, however, is a very different matter.

    Of course, in an ideal world you'd have a different password for everything, each one no less than twelve characters long and made up of both lower and uppercase letters, numbers and special characters. But of course, certain compromises have to be made for the sake of your ability to remember them all; few people could reliably recall more than two or three different ones. The trick is to make them hard to guess or dictionary-attack, yet easy to remember. The registration number of your car or your postal or zip code are two good candidates, as is your social security number or its local equivalent up to a point; the latter can be used for identity fraud, but only in conjunction with your real name, as it usually needs to be backed up by photo ID. Alternatively, if you have the dubious distinction of fluency in l337-speak, it can create nearly uncrackable passwords from any word translated into it.

    Finally, if you take away nothing else from this tutorial, remember this; a strong password is the beginning of good computer security, not the end. Never let anyone you don't know use your home computer unsupervised, or your work computer at all, and shield your keyboard if you input them on a PC in a public place like the library or an Internet cafe. If you instruct Firefox to save your passwords, use the Master Password function and set it with the strongest one you can think of. Pay attention to the address bar when following links in emails, and if it claims to be sending data over a secure connection, look at the little status bar along the bottom of the browser window; unless you see a little padlock icon, it's not as secure as they want you to think. If you keep scanned copies of debit or credit card receipts, bank statements or anything else with the card details on, black out the card number and security code with a felt pen before scanning them.
    Last edited by Jake; 28th Feb 09 at 9:40 AM.

  2. #2
    Banned BmB's Avatar
    Join Date
    Apr 2004
    Location
    Denmark
    A more audacious tactic is to steal or fake up an ID card and simply brazen your way into the building, and then out again whilst carrying the file server!
    I'm not sure how plausible this one is.

  3. #3
    Member Jake's Avatar
    Join Date
    Nov 2008
    Location
    Dorset UK
    Mike Meyers claims to have brought it off once, when challenged by an acquaintance to defeat his new security setup.

  4. #4
    Member Moleman's Avatar
    Join Date
    Jan 2008
    Location
    Arizona
    A more audacious tactic is to steal or fake up an ID card and simply brazen your way into the building, and then out again whilst carrying the file server!
    i could belive this one, in my old highschool, a kid came in with a jump suit on and walked out of the school with most of the server rack's on a dolly, about 15 min later after he left, an announcment went up for a reward for the stollen tech XP

  5. Child's Play Donor Technical Help Senior Member General Discussions Senior Member Boardwars Senior Member  #5
    Gimme your lunch Moeney! Moe's Avatar
    Join Date
    Oct 2003
    Location
    #homeworld
    An interesting thing to note is that with the availability of cheap HDDs with massive amounts of storage space password cracking has become easier. By using rainbow tables you can trade space for processing time.

  6. #6
    Member Moleman's Avatar
    Join Date
    Jan 2008
    Location
    Arizona
    moe, would you be able to elaborate one what Rainbow Tables are? this is the first time ive heard this term

  7. Child's Play Donor Technical Help Senior Member General Discussions Senior Member Boardwars Senior Member  #7
    Gimme your lunch Moeney! Moe's Avatar
    Join Date
    Oct 2003
    Location
    #homeworld
    They're a tradeoff - HDD space for CPU time, basically. Say you want to crack a hash. If you brute-force it, it'll take loads of time because your CPU has to calculate the hash for each input. Now if someone does this and saves all of his results in a table and makes that table available, the next guy doesn't have to re-calculate the hashes. Instead, he can just look them up, which is less CPU-intensive.

    Take these forums for example. Let's say your username is Joe and your password is abcdefz. That password isn't stored as clear-text, instead it's stored as a hash. There are different hashing algorithms, but they all have one thing in common - they can't be reversed. An example of this would be a very simple algorithm that assigns a number to each letter of the alphabet and adds them together. So, for our example that wold mean (using a=1, b=2 etc):
    1+2+3+4+5+6+26 = 47. This is the hash for your password. In other words, somewhere in the relicnews database there's a table containing user passwords, and the entry for Joe would be "47".

    Anyway, if I gained access to the database and looked up your password hash (47) and wanted to find out what your actual clear-text password is, I'd have to brute-force it. So, I'd start with "a" and calculate the hash, which would be "1". Then I go to "b", and so forth. So far this is easy, but at some point I run out of letters and have to add characters, so now I have to calculate "1+1" and so on. Obviously real hashes are a lot more complicated, making the calculations more demanding. With a rainbow table, I have that information right in there. Instead of doing 1+1 I can just look up the hash for "aa" in the table.

    Table sizes vary according to type - alphanumeric, special characters, etc. They're typically several GB in size, with the more complete ones that have all unicode chars being several dozen or even hundreds of GB large.

  8. #8
    Member Jake's Avatar
    Join Date
    Nov 2008
    Location
    Dorset UK
    I set out to write a tutorial and end up learning something new myself. Go figure.

    So what's to stop one from decompiling the database file, or the forum copy of SQL Server, and finding out which hashing algorithm it's using from the source code?

  9. Child's Play Donor Technical Help Senior Member General Discussions Senior Member Boardwars Senior Member  #9
    Gimme your lunch Moeney! Moe's Avatar
    Join Date
    Oct 2003
    Location
    #homeworld
    Even if you know the algorithm (which isn't difficult, it's probably a well-known hash algorithm that's used everywhere), you can't reverse it. The only thing you can do is crack them, i.e. try out various clear-text passwords, run them through the algorithm, and see if the result matches the hash.

    There are mathematical proofs demonstrating why you can't reverse them, but the very simple answer is "by hashing you lose information. You cannot regain lost information". Here's a little example:

    I tell you a number - 47 - and that I got that number by adding various numbers together. You'll get $10 if you guess which numbers I used. Problem is, there are quite a few ways to do this, ranging from adding 47 ones to adding 23 and 24. In other words, you can't recreate my original chain of numbers even if you know my algorithm ("add them all up") and the result ("47").

    But since we don't want someone to even try that, we don't exactly make the SQL database available to the general public.

  10. Forum Subscriber  #10
    Logico-Fishosophicus ionfish's Avatar
    Join Date
    Mar 2000
    Location
    I live in Beer.
    And this is why we have salts, and other key strengthening techniques.

  11. #11
    Banned BmB's Avatar
    Join Date
    Apr 2004
    Location
    Denmark
    Um, if the server verifies by hash, isn't it as simple as giving any password that results in 47?

  12. Child's Play Donor Technical Help Senior Member General Discussions Senior Member Boardwars Senior Member  #12
    Gimme your lunch Moeney! Moe's Avatar
    Join Date
    Oct 2003
    Location
    #homeworld
    Theoretically yes, but obviously hashes don't work by simply adding up numbers. If you have the hash however, what you can simply do is write your own login cookie and skip the whole password thing. What I described assumes that you want to get the clear-text password.

  13. #13
    Member basuatreya's Avatar
    Join Date
    Sep 2004
    Location
    Toronto, Canada
    This reminds me of something that happened recently. A customer of ours at work had a falling out with one of their admins and they couldn't perform administrative tasks for our software.

    It was a big problem because this customer was quite important for us, but more importantly we are in trouble if we can reverse the security of our own software, but if we can't resolve this problem the customer was threatening to cancel our support contract.

    We decided to help break our own security, so I'm looking at all of the hashes and they all look familiar. So I'm thinking the customer hasn't changed their passwords from when the software was originally installed. But then when I ask the customer to show us what they are doing, it turns out they do have all the admin privileges, they just don't know how to use the software.

    We dodged a bullet that day.

Thread Information

Users Browsing this Thread

There are currently 1 users browsing this thread. (0 members and 1 guests)

     

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •