Might be time to change your passwords

A common encryption tool used to safeguard transmitted data has been breached.

http://heartbleed.com/

http://www.usatoday.com/story/tech/2014/04/09/heartbleed-five-questions/7501033/

Some websites have implemented the fix, but not all. As I understand it (I'm not a systems programer); If the site is still vulnerable, all changing you passwords will do is allow an eavesdropper to capture your new password. Not vulnerable sites have been fixed and should have the passwords changed, and No SSL sites require no action. Though if you share a password/username with an affected site, it would be prudent to change it.

https://github.com/musalbas/heartbleed-masstest/blob/master/top1000.txt

The breach has been fixed, but servers all over the world must install the latest versions of the OpenSSL software.

The breach can be compared to having top-notch security all over and around your house -- but if someone goes to the exact right location behind your garden shed and peeks into the gutter he can extract random pieces of information from the internal workings of the lock to your front door. Repeat this many times over and he will have enough information to replicate the keys to your house (and the code to deactivate your alarm.)

No doubt criminals all over the world are rushing to take advantage of this security flaw before all servers are upgraded. The smaller organisations with less resources will be the slowest to upgrade, and thus vulnerable for a longer time. Anyone seriously about internet security should have upgraded openSSL yesterday.

I've heard estimates that OpenSSL would be running on something like 60% of the servers on the internet.

What no one knows is: Has this flaw been known to criminals before the day before yesterday? If so, they've had ample time to snoop around and sniff out vital security information. If NO ONE with bad intend didn't know about this before... then we've been very lucky, to say the least. But even _*if*_ we are so lucky, criminals will surely be working like mad right now to build tools to extract as much as they can from the remaining servers that hasn't upgraded their openSSL software.

Yes, change your passwords now, especially for the biggies (bank, eBay, credit card, online trading account, anything that can cost you serious money or trouble, or delete important data, or facilitate identity theft). And probably change those passwords often until the dust settles.

This one is big. As an example, the Canada Revenue Agency (equivalent of the IRS in the US) shut down all online access this morning as a precaution. Three weeks before tax returns for the whole country are due.

It's unbelievable, disgusting, and grossly negligent that a gaping hole like this would be "in the wild" for two years.

Your NSA tax dollars at work. The Cyber warfare must be getting pretty serious with the Russians for the NSA to throw away their ability to crack SSL in the last 2 years now that the Russians can exploit the same engineered vulnerabilities.

Windows has been in the wild a lot longer. Since 1985.

It is not easy to exploit this OpenSSL bug. In fact, I don't think there are any reported cases of it ever having been exploited. It's been fixed for a while already. I believe immediately after it was fist found, it was fixed, and new software was available for immediate installation.

I am a LOT more worried about some hacker getting into a merchants database and stealing data in bulk. That method is preferred by hackers, because they get so much more data for their efforts. They aren't going for the "small potatoes" of trying to intercept individual transactions going across the internet so much anymore. The NSA is doing that, but not the hackers trying to steal your information.

The following is a public service announcement from your local neighborhood network security engineer.

Never, ever, re-use passwords across multiple services. If you use the same password at your bank that you do anywhere else, once that password is compromised all the associated services will be compromised.

Some vulnerabilities can be mitigated by using strong passwords. A strong password is at least 12 characters long, contains at least one of each type of character (upper and lower case letters, digits and symbols), and makes no word in English or any other language. I use a secure password database and randomly generated complex passwords. If you must come up with a memorable strong password, create a sentence such as: "Doug Ritter runs the best forum on the Internet, especially for a knife guy." That sentence could be the mnemonic for a password like "DRrt^f0tI,efakg"

Other vulnerabilities, such as Heartbleed, cannot be mitigated by using strong passwords. That's why it's so important to use different passwords everywhere.

One more thing about Heartbleed. If you run a service affected by it, replace your SSL certs and keys after patching the vulnerability. If a service you use is affected by it, change your passwords after the service you use patches the vulnerability.

I'm afraid you're incorrect. This vulnerability is being exploited in the wild. In addition to other reports one of my customers was hit by it. Also, the fix for this was released on April 7th.

Unfortunately OpenSSL is a library, not a full product. What that means is that many, many products that use OpenSSL must be patched to use an unaffected version before the problem is truly solved.

Everything passing across an SSL/TLS link connected to an affected service should be considered compromised, including user credentials.

It's about time for me to get off my ass and finally implement better password security. I have about 100 computer-related logins that require a password. So, I need to think of a system that doesn't repeat a password but allows me to memorize at the same time. This is a daunting task.

That worked for me in the early '90s. You need to get yourself an encrypted password database.

Here is how I do my passwords:

I start out with something that I have bought recently that is easy to remember:


I rearrange that a little, still easy to remember (notice everything is lowercase now):


I replace letters/numbers with their "equivalent". e.g., e with 3, s with 5, a with 8, l with 1, o with 0, etc. I do it both ways ... so e with 3 and 3 with e.


Then I alternate holding the <shift> key down to capitalize every other keystroke:


I look at the end result and if it doesn't look random enough, or doesn't end up with upper and lower case letters, digits, and punctuation, I buy something else (yeah!) and start over.

I end up with very strong passwords this way. And they're easy to remember. I should say, the passwords themselves are NOT easy to remember, but the sequence of steps to generate the password from my example seed phrase ".357rugermag" is easy to remember. And the seed phrase is itself easy to remember because it represents some cool item that I recently bought for myself.

The downside to this is that I cannot tell anyone else my password. I cannot even sit down and write it on a piece of paper. I have to have a normal QWERTY keyboard in front of me so I can visually see things as I hunt-and-peck the keys while alternating "shift key up, shift key down, shift key up..."

This is how I do the passwords for stuff that I really need to be secure. But for passwords for the less critical stuff, say for my login here on ETS, I use simpler passwords. I have lots and lots of these less secure, but still decent quality, passwords. Since I can't remember them all in my head, I store them in the "KeePass" application. I have that for Linux, my Android phone, and Windows. I assume KeePass might be available for iPhone and MAC's too, but I don't know for sure. The encrypted database for KeePass is copied transparently between all my devices.

Each shopping_website/bank/etc. that needs to be secure has its own password - they are never the same password shared between sites. But I will admit, for some of the internet forums I visit, like ETS, I occasionally use the same password. That is because the ramifications of somebody hacking my ETS forum account are pretty minor.

I hadn't heard that. Now I have. Thanks.


I remember reading about this bug several weeks ago (a month or more maybe?). And the fix was reportedly available back them. Possibly the fix had been completed, but not made generally available yet. I don't remember the exact details.

Which password are best depends on how you expect it to be hacked. Do you expect a person to randomly guessing a password or a brute force attack.

For a brute force, generally a very long password is good, not necessarily one with complicated letters/signs.

I have separate passwords for impotent things and a few I use on non essential things like forums.

Are you sure you're thinking of Heartbleed? The GnuTLS vulnerability was reported a month ago and mentioned here on ETS.

Schozzbott!

This article reports on the German developer who introduced the vulnerability into OpenSSL two years ago. A trivial coding oversight that has massive potential ramifications.

Password strength:

This will evaluate passwords for strength. The results are surpising, some passwords I though were "best" only came back "strong" and vice versa

https://www.microsoft.com/en-gb/security/pc-security/password-checker.aspx

Wikipedia has a pretty good article on password strength
http://en.wikipedia.org/wiki/Password_strength#Entropy_as_a_measure_of_password_strength

And just for chuckles, the 25 most common, and worst passwords used in the last year.
http://www.slate.com/blogs/future_tense/..._about_you.html

Yeah, who would have thought "password" was "medium" and "password123456" was "strong"? Thanks for verifying that, Microsoft!

"aaaaaaaa" also is medium, while "aaaaaaaaaaaaaa" is strong.

Looks like all that stupid Microsoft is doing with that password checker is counting the number of characters. 8 characters - no matter what they are - are considered medium, 14 characters, strong. How brain-damaged is that? But it pretty much illustrates the security model that Microsoft uses in Windows, and explains the rampant security issues constantly being discovered there.

Try your passwords here https://www.grc.com/haystack.htm
Adding more characters, even if they're the same does appear to make it more difficult to brute-force crack.

Linux is not much better or could be worse. Running password12345 through cracklib-check gives it the "ok" as a password. Yet password123456 or password1234567 gives the result of too simplistic/systematic.

Similarly, 123aaabbb gets the ok from cracklib-check, but 1234aaabbb is too simplistic/systematic.

The design of cracklib-check heuristics can give a false sense of security to the user just like other password checkers/validation programs.

No matter the password strength, I check my bank accounts online once or twice a day. 24 hours will not pass before I can respond to foul play. I think that's better than having a super strong password and checking bank accounts only once a week.

By the way, I do need a system that allows me to memorize, even though I will also keep a password vault program. Every now and then I'll find myself somewhere with nothing but an Internet connection on a random computer. I need my password to be in my brain at that moment. Thus, I need to memorize at least the important passwords.

Typing your important passwords into a "random computer" might be something you want to re-think.

I would use a random computer for an emergency. Imagine something like a foreign country where the only/best means of communication is the Internet.

Anyway, I am "the guy" in the family who must know passwords, not only my passwords. For example, over the past ten years at least ten times, I've received a call from a house security company who was about to send the cops over to one of my relatives' house. I knew the password every time and prevented the cops from showing up. (I knew they were not intrusions.) I have been the only one who ever remembered the passwords because of my system for recalling. And, by the way, around me, the cops have the right to bust into the home if they get a call from a security company. That can easily end all bad.

So, about knowing passwords, I'm talking about the entire universe of passwords (including house security, padlocks, safes, etc) that I like to keep in my brain for quick recall. The security vault software is a backup.

This is an outstanding cartoon that I saw awhile ago. Since I have not heard anybody refute this method, I will be implementing this.

I love XKCD but I do not entirely agree with the "correcthorsebatterystaple" password creation methodology.

For straight-up unsophisticated brute-force attacks, where every possible password is tried starting at "a" and ending at "zzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzz" -- the esteemed Mr. Munroe is correct.

Most password attacks are far more sophisticated, using dictionaries, precomputed hash tables, and so on. A dictionary attack against "correcthorsebatterystaple" would succeed far more quickly than a brute force attack.

A dictionary attack to get "correcthorsebatterystaple" would consider about 200,000^4 combinations of words. How long would that take? It would actually be more combinations than that because the hack doesn't know how many words to consider (e.g., 1 word or 9 words?)

By the way, I don't think it's possible to do a pure dictionary attack on "!correcthorsebatterystaple" (one random symbol in there).

With the cost of hardware getting cheaper all the time, I wouldn't be surprised if an individual COULD create a hash table in RAM that contains a password like "!correcthorsebatterystaple". I'm curious if cloud computing providers like Amazon look for activity like this by their own users? A hacker doesn't even need to buy their own hardware--just rent time on someone else's shiny, high power data center.

Of course, we're talking about the scenario where someone has unlimited ability to try and get a correct password, like stealing the password file or password hash table for some site and running an attack against that file. But if someone can access THAT file, that system already is already in deep trouble.

It's like not someone can try a million passwords by trying to log into your back account through the login webpage. I can only try to log into my work PC three times before it locks me out for 6 hours or until I call an admin and they deactivate the lockout. In that kind of environment, you don't need military grade passwords.

It's old school (practically ancient, in Internet time) but I'm still a big fan of the diceware method, which basically gives you "correcthorsebatterystaple" type passwords. I almost always have at least one number in it, so a simple low security password would have the format word + number + word. Depending on how strong I want the password to be, I will also use diceware to randomly insert/replace a letter with a special character as well as add more words.

A much better system than random gobbledygook passwords, as long as you remember that the strength comes from the length, not the apparent "randomness" of the letters and numbers.

All passwords are a compromise between convenience and security.

My previous WiFi password was (cut-n-pasted directly from my archives of old passwords):


I finally ended up replacing that with a simpler one, because nobody could ever type it correctly to attach to my WiFi network. Including me. After about four tries, everybody gave up except for the masochists. It got to be too much of a pain-in-the-butt when I was configuring devices like my Roku video streamer, which don't allow for copy/paste from an attached thumbdrive containing the password. There's quite a bit of "is that an ell or an eye or an oh or a one?" in there, depending on the font I used to print it out. It was, however, quite effective at squelching most folks desire to even attempt to connect to my WiFi!

In this case, I think I had a very "secure" password, but certainly not a "convenient" one. It was overkill for the task at hand, especially seeing as how my WiFi signal doesn't even make it out to my property line.

I have a similar password for my Wifi. I think Comcast started doing this so there is zero chance a random person is stealing bandwidth. lol

Speaking of "stealing bandwidth," I briefly skimmed an article a while back and never was able to go back and read it more carefully, but I was surprised to learn that at least one major Internet provider (and I can't remember which one) allows any customer to use any other customer's home wifi by logging in with their own customer login info. Has anyone else heard of this? I wonder if these customers even realize that they are potentially sharing their wireless access point with complete strangers?

Apparently the public bandwidth is separate and will supposedly never degrade your own bandwidth or allow someone from outside to access your private network, but still, I was pretty shocked to hear about this. I have AT&T and I remember feeling relieved, so I don't think it was AT&T that has this "service".

http://www.foxnews.com/tech/2014/04/11/h...b-use-dhs-says/

DHS is saying that there have not been any reported attacks or incidents involving Heartbleed. So I wonder what the real truth is, has this bug been exploited or not? Not that I trust anything DHS would say all that much...

Nobody knows: Exploiting the vulnerability leaves no trail in the system being tapped.

I have a non-zero number of customers who were attacked. There are public reports of attacks at http://www.theregister.co.uk/2014/04/11/hackers_hammering_heartbleed/, http://news.yahoo.com/u-government-warns-potential-attacks-heartbleed-bug-135137709--sector.html and other places.

The exploit itself does leave tracks, depending on the service's logs. Services using the OpenSSL libraries can be configured to not log the right information, however.

Again, this is not always true, depending on the sophistication of the attack.

The only sane approach is a password manager. I use KeePass. This will generate new passwords of whatever length or alphabet you want, and keep them in a database. The database is encrypted, so you can back it up to non-secure locations. I use DropBox as an off-site backup and as a way of replicating the database to a variety of devices - desktop, tablet, phone. I have to remember the KeePass master password (which is a long, nonsensical phrase), and the DropBox password.

You can get add-ins for browsers that will attempt to recognise web pages and enter the correct password for you. I found they weren't reliable enough, and they also mean having the database open all the time you are browsing, so now I just copy and past between KeePass and the browser as needed. No-one shares my machine so I don't mind leaving websites logged in, so I don't need passwords every time.

I actually keep two password databases. The second one makes low security passwords more convenient. It has an easier master password that I can type quickly, and I don't mind leaving it open for extended periods. I use it for websites that don't have much at stake, especially forums.

There are several other password managers. I like KeePass because it is open source, and stores its password database locally. Some others store their database online, which means you can get to it from any device, but I think means you have to trust them more. Whatever you use, it should give you strong passwords that you don't need to memorise, and it avoids you ever having to reuse passwords. Just make sure you don't lose that password database or forget the master password.

The answer is in the picture. It is not 200,000^4 combinations because those words are so common. If you look at his numbers, he is only claiming 11 bits of randomness per word, which means a dictionary of about 2000 words. 44 bits of randomness altogether. 550 years at 1000 guesses per second. In practice they can be a million times faster. It's only 5 hours at a billion guesses a second, and if they have a big cluster of GPUs or a botnet they could be hundreds of times faster than that.

Upshot is that 4 random words, 44 bits, isn't enough nowadays. It's better than Tr0ub4dor&3, but that's not saying much.

That doesn't make as much difference as you might expect. Checking all one word passwords, then all two word, then all three word, doesn't take much longer than checking all three word passwords because there are 2000 times as many three word passwords as two word ones.

Those kinds of transformations are known to hackers and easy to automate. There's a good (if long) article about hacking that kind of rule-based password on Ars Technica.

Since you are using KeePass, why don't you let it generate strong passwords for you? Your ".es7rug3rm8g" is 75 bits, which is much better than "correcthorsebatterystaple", but KeePass routinely gives me over 128 bits.

Apparently, the NSA knew about Heartbleed bug and took advantage of it, but the NSA denies the charge.

http://gigaom.com/2014/04/11/nsa-knew-about-devastating-heartbleed-bug-and-used-it/

[DELETE]

Which is why I use the "every other character with the <shift> key" part. Those digits, which I agree are easy transformations to automate, become punctuation characters when you use the <shift> key. Does this actually make the final result password more secure? I can't say with any certainly because I haven't done any personal crypto testing myself, but at least we can probably assume it doesn't make them any LESS secure.


I have thought about that. If I only used my desktop computer to access forums and such (where I tend to use the less secure passwords) I would do this. However, KeePass on my Android phone is not as easy to use, nor is the cut-n-paste, compared to a desktop computer. So I just have not made the switch to using KeePass to generate good passwords. I justify this, rightly or wrongly, with my thought process that "I don't really care all that much if my forum passwords get hacked". The consequences for me are minimal, since I don't share those less secure passwords with bank websites and such. Still, I agree it would be better to have super-secure passwords everywhere, no exceptions. I just haven't made that move yet. I should.

I saw something, somewhere, on another site...where a mathematician showed the compared probabilities associated with "m4Nuf4C+ur3D" passwords and those associated with passwords made of four 'random' words, for which a mnemonic could be contrived. The four random words were much more secure: e.g. Looking at my Avatar, one might pick "CrackerPonyFlipFlop". This kind of password is easier to remember AND more secure...and I guess I won't be using "CrackerPonyFlipFlop" anytime soon...

Mapping digits to punctuation via the <shift> key is also easy to automate.

It's probably not less secure than ".357rugermag", but that's only about 45 bits so it's not a very high bar for a super-secure password. It's probably not as secure as 12 genuinely random characters would be (84 bits). Whether it's secure enough is a judgement call. It may also be something which would likely remain unbroken today, but become weak as hackers catch on to those transformations.

Fair point. I don't use many passwords from my phone. There's always a trade-off between security and convenience.

Thanks, KeePass 2 is way better than SplashID.

This isn't secure at all. A dictionary attack will find all 4 words just as easy as it will find one.

Basically you take the first letter C and go down in your dictionary file to all the words that start with C. Then look for all worlds that are Cr within that subset, then Cra, etc.
Once you've found Cracker you just take the next letter P and look for all the P words.

I'll give an example of how easy it is to do: Way back in my high school English class we have to figure out compound words by breaking them apart into the Greek and Latin roots, then getting the meaning of each part and putting those back together into a definition.
I did this in Commodore Basic in 1989/1990.

That's still not a trivial computation. The computer must be prepared to try every combination of known words. If one character or one capitalization is out of place, then the dictionary attack won't work.

Anyway, obviously, the most secure passwords are going to be long and random. Any rule that deviates from that can be met with at least some criticism. There is a point at which your password is no longer the weak link (e.g., it may be much easier for a hacker to steal your computer's key strokes).

With a few simple modifications to the cartoon "CorrectHorseBatteryStaple", you can obtain a password that makes a dictionary attack substantially more difficult without being much harder to remember.

Here are some examples:
CorrecHorsBatterStapl (Drop the last letter of each word)
C!orrectHorseBatteryStaple (One random character in there)
cOrrecthOrsebAtterysTaple (capitalize second letter of each word)

It's not always possible/convenient to have a "long and random" password. Thus, some sort of mental hash algorithm is often required.

Notebook, my friend. Notebook.

Yeah, yeah, don't write down your passwords. What are the chances that the guy hacking all the world's servers just happens to be burgle your house?

It's easier for me to memorize patterns and pictures. For passwords that I want to be highly secure and memorized, I use a pattern on the keyboard. I don't even know what the password is because it's gobbledygook, but I know the pattern/drawing in my head. The password can be quite long and is not written down anywhere. I do write down something that can make me, and only me, recall the pattern that is meaningful only to me.

So, even if a thief hacks my password vault, he is still out of luck. A brute force attack is the only crime that will work. The only downside to this approach is that I must have a full QWERTY keyboard to input the password.



Good old pen and paper!

Thats why I was just illustrating how simple it is. You don't need to look for combinations of words since you start comparing from the beginning of the text string and once you match a word or words you match the next. In your example it would just take 4 passes.

A dictionary of millions words (and misspellings of words) is actually quite small and easy to get ftp://ftp.gnu.org/gnu/aspell/dict/0index.html

So even intentional misspellings are very trivial, a non programmer like myself can throw together code to do it.

Adding other characters does make a simple attack harder, but even then the common h@ck3r$p3@k is well known so you simply run a script against your dictionary to find each word with a commonly substituted character and append those. You use a dictionary like I listed above which has common mis-spellings of words. Then if your intentional addition of a character doesn't match the spell check or hackerspeak dictionary you just run a different attack vector.
This can all be done in seconds, remember modern database servers are doing millions of lookups like this all day long when you check your back account status for example.

Anyway, don't get a false sense of security thinking 4 words (or 6 or 8 or 10) is more secure, it just makes the attack take 4 (or 6 or 8 or 10) seconds rather than 1. The 4 random words is a tiny bit more secure than one but not "much more secure" as stated originally.

This is really something I don't understand; I know in the movies they show passwords being decrypted this way, one character (or as you are describing, word) being decrypted at a time but I always assumed that like pretty much anything else shown in the movies with computers, that this was yet another misrepresentation. Movie makers tend to think computers are magic.

My assumption was that a brute force password guessing algorithm would basically only get a boolean result; it worked or it didn't. I don't understand how it could know it was partially correct unless somehow it had access to the encrypted password, but then I would think that you'd be dealing with an entirely different type of algorithm.

But again, this out of my area of expertise. Any references that would help me understand this better?

That's what I'm thinking also. I don't understand how a computer can know if a password is partially correct. If that were true, a brute force attack on the following password could be cracked rather easily on the weakest of computers:

&*&*(89234897sdlkjehruipIIOPUE3R-708760340=-23-0-0978s-89234^&^234897^&^7@3

In other words, the computer would be able to know the correct first character and then move on to the next, etc. I'm almost certain it is NOT possible to crack a password this way.

By the way, again, it's always possible to criticize anything that is not "long and random". It would be helpful to propose an improvement on a password recall system if you see a notable weakness. Not everybody can have their password vault program/notebook with them at all times.

It depends on the type of attack.

If the attacker has extracted a hashed password table, a brute-force attack can run every combination of letters and numbers for a single password out to twelve or more characters lightning fast. Using this (dumb) method I could typically crack 500 passwords in an hour or two on my (underpowered) laptop. Using an AWS cluster or something similar, we could get the same job done in a second or two. This was legal, by the way, because I had written authorization from my customers to do it.

Using a good dictionary, we could cut the time by a factor of roughly ten -- instead of 120 minutes, maybe 12. "Correcthorsebatterystaple" passwords would fall quickly to this method. But we'd be back to the brute-force type of attack for a strong password like those I mentioned upthread.

Using rainbow hash tables, if we got hits we'd get the whole thing done in a few seconds.

If you're not working on extracted password hashes, but rather attacking via a user interface, things get much, much slower.

If there is access to the hashed password table, that's a completely different ballgame. As far as I'm concerned, that falls under the umbrella of "stealing a person's password". All bets are off at that point. It really doesn't matter how long or how random your password is.

I'm oversimplifying a little to keep from wiring a whole book

basically a dictionary attach we take a bunch of known words, we just break the password down smaller and attach smaller words. Say a 9 character password I search my dictionary I check for combinations of smaller words such as appletree, brownpony, applepony, browntree as well a 9 character words. Then since I've used a spellchecker dictionary I have the common misspellings so we check brownpnoy, applepnoy, etc. We also added the hacker speak to our dictionary so we look for br0wnp0ny, @pplep0ny, and of course we did the search replace against our whole spell checker dictionary so we got @ppl3pn0yand br0wnpn0y as well.

Basically I am disagreeing with the statement that simply making a password from four (or most any number) words make it much more secure because it does not. Adding random letters, numbers, punctuation, etc does make it a little more secure.

Usually when I need a password I'll go find a random password generator and I'll have it generate say 10 at a time then choose one or more and merge them. That way even if someone were sniffing the traffic at that instant they don't know which of the 10 I chose.

Why wouldn't you use a random password generator that is disconnected from the grid?

Non admin on work computer so I can't install anything there. Passwords at home I can do that.

I use an encrypted password database app that syncs with the desktop version of the same app on my home PC. It includes a random password generator, and I use it constantly.

The vast majority of password thefts come from stealing a hashed password table or compromising a workstation and either stealing passwords stored in a browser or sniffing keystrokes.

No amount of password complexity can help against any of those scenarios when faced with a determined attacker. However, most attackers aren't determined enough to crack long complex random passwords. If nothing else, such a password may give you time to find out about the breach and change your password before it's cracked.

For password attacks against user interfaces, long complex passwords are very effective.

Here's an interesting article that cleared some things up for me: Anatomy of a hack: How crackers ransack passwords like “qeadzcwrsfxv1331”

Based on the little reading I've done so far, it seems like the biggest risk is having the hashed passwords stolen from a compromised website which then allows the crackers to decrypt the passwords at their leisure.

The real test is to download and learn to use the password cracking tools. Then see how long it takes to crack the password your using.
Just be careful searching for those tools, don't use an unsecure browser like IE unless your doing it in a sandboxed VM as those sites also like to test your browser security

Scary numbers. Are we talking cracking a salted password hash table, Chaos? If you're saying you can pre-compute a rainbow hash table for a single salted password like "Correcthorsebatterystaple" and get a hit in a few seconds, that kind of speed would blow me away. And would make me think about taking out all of my money from the bank and burying it in a hole in the woods!

On a related tangent, if someone has access to the password hash table on some server, the user is already in deep trouble. Which is why you shouldn't make the bad guys' lives easier by reusing passwords (or usernames) for important websites/accounts. That is, a bad guy has already compromised the server for system A if they can grab the password hash table. Don't make it easier for them to get into your account on system B by using the same username/password from system A on system B.

And also why having the ability to use unique email addresses for each website is worthwhile to me. Many websites use an email address as the user account value. If a hacker can obtain the username, email address and password for me from system A, that info will not match on system B, C, D, etc.

For example, I've been a longtime Yahoo email user. The paid version allows you to create unique passwords in the form of rootword + whatever @yahoo.com. That way, you can use a unique password for each account and make life tougher for the bad guys if the email address is the username for an account.

If they can't just reuse your online poker username/password at the login screen for your bank account, then they'd have to try some other method--actually hacking into the bank's server, using a "spear phishing" attack against you, keystroke logger, etc. Except for the spear phishing attack, that's a lot more work to get at your paltry bank account balance and probably isn't going to happen. Then again, my money might already be in that hole in the woods by then!

Equally useful is that this email feature lets you more easily cutoff spammers by deactivating certain email addresses without having to totally throw away an email account and go through the trouble of changing the email address for all of your accounts. Although, I have to admit that spam filters work remarkably well nowadays, so I haven't had to deactivate an email address in quite a long time.

The way that passwords are typical stored, there's no such thing as a partially correct value.

Passwords are never (should never be) stored as-is. You run them through an algorithm (a one-way cryptographic hash) which spits out a long gobbledygook string, which looks like c11083b4b0a7743af. This string is what is actually stored, not your password.

When a bad guy is trying to crack passwords, they also have to hash their guesses and then compare that result to the result in the password table. When they match, they know they have guessed the password.

There are two main characteristics of one-way hashes that are useful in this case. "One-way" means that you can't start with c11083b4b0a7743af and back-calculate what the password is. That's why it is safe to store the hashed value rather than the plain text password.

The other important feature--and this answers your question--is that even the smallest difference between two passwords should result in very different hashes, so even a password off by the last letter will have a totally different hashed value.

Therefore, it's not possible to know if you have a partially correct password guess. You either know the whole password or you have no clue how close you are.

Yes.

Burying money in a hole in the woods is probably not as safe as keeping in an insured bank or credit union, though.



Exactly.



The most common ways an attacker can steal money from your account go like this:

• Using a card and keypad skimmer (or card skimmer and camera) to get your card details and create a forged card to extract money from your account(s)
• Extracting payment card details from a retailer you do business with, either where the data is stored or while it's in motion
• Compromising the credentials you use for online banking and creating fraudulent transactions

Addressing these threats, here is what I do and what I recommend:

• Do not use or possess a debit card. When a debit card is compromised, the money is gone and under the law you are guilty until proven innocent. When my financial institution insisted on replacing my ATM card with a debit card, I insisted in turn that they completely disable the debit function. If you absolutely must use a debit card, keep a separate account for it and only keep money in it that you don't need to pay your mortgage and other bills. But you're better off without one altogether.
• Use an ATM that's less likely to have been compromised, such as one at a branch of your financial institution.
• If you're paying with some means other than cash, you really don't have any control over whether the retailer will adequately protect your payment data. So use a credit card. You're protected by law with a maximum liability of $50 (most cards have written policies of $0 liability) and the stolen money isn't removed from your account keeping your bills from getting paid.
• Use good antivirus software, keep your operating system and browser updated, and don't surf anywhere that might be a bad "neighborhood" on the Internet. Stronger security might be had from "walled garden" operating systems like iOS on the iPad, and there's a good argument that Macs and Linux is more secure than Windows.
• Consider using a dedicated workstation that does four things: OS updates, browser updates, antivirus updates, and online banking. I don't do that, but I'm also very cautious about where I surf.
• Delete all spam unread, never click on a link in spam email, never open an unexpected email attachment.

That's true...but unfortunately incomplete. Hashing algorithms are imperfect, and every one of them have "hash collisions." This is what happens when two inputs into the hash algorithm result in the same hash. Which means that there may be more than one password that can open your account.

I don't know about the debit card thing, at least from a Canadian perspective. The debit card is pretty much the defacto method of payment up here (over 10 years ago debit outpaced cash for purchases). However, there have been several changes over the years to improve security both from a technology perspective (chip cards) and process perspective (we usually insert/swipe our own cards).

However, we also have some protection against fraud, similar to credit cards. Here's some good info: Debit Card Fraud

Its also way nicer to spend the money you actually have

A big +1 on the do not use or have a Debit card. Chaosmagnet suggests that if you do, have a separate account for it. BUT be careful that the bank does not have the right to tap into other accounts if the Debit card is overdrawn: I think most Banks now are required to offer the option to not overdraw the debit card account, but, as I understand it, you must opt into this option. We do not even have a debit card because of the added risks and liability mentioned. Our solution is a credit card, which is payed off monthly.

Also, ATM machines at bank branches are not more secure. 1.) due to their heavy use, they are a prime target. 2) They get lots of traffic when the bank is closed (and no one is there to keep an eye on the machine). This happened at my bank branch, and it hit a lot of people. There was some theorizing that the skimmer was being installed after the branch closed, and de-installed before it opened. Thus, the branch employees never noticed anything different about the machine.

It would not be good if these clients of yours are still using MD5 for their passwords since it has well known collision problems. I thought more recent hash functions like the SHA-2 class of hash functions basically eliminated collisions.

I generally agree with what Chaosmagnet said about bank card practices.

I've had fraud on my credit card a few times. It's a non-stressful thing because 24 hours does not pass before I check my bank accounts and credit cards at least once. The bank immediately reversed the charges and started an investigation. Had the fraud been on my ATM/debit card, I would have been stressed.

I don't use my debit card for anything but the ATM. You guys have reminded me about fraud that occur on a debit card, and I will see if I can get that feature removed. By the way, I have never come across a vendor who accepts debit card but does not accept credit card. So, I use credit card everywhere I can. Further, I can't recall ever using debit card for anything my whole life.

Another good thing about using credit card is that if you time your payments properly, your credit score will keep rising. My lowest credit score is over 800, and I think it's due to how I use credit cards: I let the card show a balance on the statement date, and then I pay it off.

Most of the skimmer attacks I've read about involved off-site ATMs. As you say, that doesn't mean that branch ATMs are necessarily secure.

Fewer, not eliminated.

Okay, when talking about password length/complexity requirements here the focus seems to be website accounts as these appear to be the most likely source of compromise. Am I correct in assuming that the same level of length/complexity would not be required for things like Windows/Active Directory accounts as they are (I'm guessing here) not likely to be open to this type of compromise. What about things like Wi-Fi passwords?

Active Directory passwords should have a high level of complexity if you care about what your password is protecting. There are several ways to extract the hashed passwords from the Domain Controller.

I spent some quality time as a wireless pentester -- it was one of the more enjoyable ways of making money I've ever found. The two most common ways of securing home networks are:

WEP -- In my opinion, WEP is worse than no security whatsoever, as it leads you into thinking that your network might be secure. In the field I have repeatedly cracked WEP keys in about two minutes.

WPA/WPA2 (PSK, Pre Shared Key, or "Personal") -- Here, password length and complexity are critically important, as WPA and WPA2 are subject to brute force attacks. Weak passwords will fall, strong passwords are unlikely to be.

For corporate networks, there are a whole mess of wireless security protocols. Some are better than others. WPA2-Enterprise uses 802.1X and is pretty hard to crack. However, unless the corporate network is running wireless intrusion prevention, there's a straightforward way of attacking it by setting up your own wireless access point and stealing authentication credentials from users.

I don't go to the VM level but the only place I use IE at all is if a site won't run without it and I HAVE to access it.

On the security side, MS has come a long way in improving IE since the IE 5 or 6 days.

I sure do have great timing. Looks like there is a very serious vulnerability in Internet Explorer versions 6 - 11 which could allow a total hijack of your PC. Sounds like this vulnerability has been exploited in the real world.

So, if you're an IE user, use something else for a while until this gets addressed. People still using WinXP are probably out of luck on getting this fixed, though, which is basically what "everyone" has been warning about as XP support ends from Microsoft. No more fixes and updates for the XP users.

Some articles mention that this is a Flash-based vulnerability, but I think they are mixing up two separate issues. More info on the IE issue can be read here .

The biggest issue with IE is the OS integration so a hole in IE becomes a hole into the OS. Any other browser isn't as tightly integrated so a hole in the browser is pretty much limited to the browser.
Thats how they can get junk through IE even when your not running as an Admin user.

Any time a relative or friend asks me to fix their system I load a new browser and hide IE otherwise they can go to someone else for support. I've seen so much get through even the new versions that its just too risky to use it.

Same with MSOE or whatever they call it now.

Speaking of which.....
http://www.usatoday.com/story/tech/2014/...ne-fox/8409857/


Original announcement:
http://www.us-cert.gov/ncas/current-acti...erability-Being
http://www.kb.cert.org/vuls/id/222929
https://technet.microsoft.com/en-US/library/security/2963983

FYI for anybody downloading Mozilla; there is a lot of extraneous programs, including malicious adware, that gets downloaded alongside the Firefox browser. You will need to go into the control panel and manually uninstall these.

BOHICA

As far as I know, if you download Mozilla Firefox from http://www.mozilla.org/en-US/firefox/new/ there is no adware or other unwanted programs downloaded with it. There may be a checkbox to uncheck for a toolbar.

Yikes. Control Panel of Windows? Or Add-Ons in Firefox?

I went to Control Panel/Programs of Windows, and I didn't see anything that looked like malware. I also went to Add-ons in Firefox, and I didn't see anything. Maybe my protection software protected me.

Mozilla does not add any software along side it, the only wan you would end up with something like that is if you download from some other place. Just download it from the real source and you'll be fine.

FYI, Microsoft issues fix.

http://news.msn.com/science-technology/m...sers-get-it-too

I'm staying with Firefox for the foreseable future with this computer. My next one will be a Linux system as MS and Google seem to be not playing very nice anymore.

The nice thing about switching to Firefox and other open source cross platform is you can run the same thing once you switch to Linux. Thats how I moved over years ago.

Seems MS had second thoughts and will issue a fix to Windows XP users after all.

If you don't like the way the current versions of Microsoft programs look (lower contrast monochrome w/ back, forward, & page reload combined with the address bar) don't let Firefox update to version 29. You will lose any custom icon locations or additions you've made.

Edit: You can get them back by reloading version 28.

I agree. I do not like the way they integrated these into the address bar. I prefer to put these icons where *I* want them, not where somebody else wants them. You used to be able to do that, but now you can't. Ironically, you could have put the icons in the same locations they are forced into now with the old version. So they didn't add anything new. They just fixed the locations for you, and took away your choice in the matter. I can't figure out what their rationale was for doing that.

You have to hunt for it but they have a page where you can download older versions. Version 28 writes over 29 or the beta for version 30 and things are back where you had them. If you use the larger icons and/or the text below them, you will have to go into customization and reset those options. I turned updates from notify me of new versions to don't even tell me about them.

P.S. I wouldn't call if fixed in anything except location. I'd call it broken. I've got home, reload, print and a couple other icons between back/forward and the address bar.

I may be the odd man out, but I like the new Firefox and will probably use it lot more then Chrome now.

For those who want to revert back to Firefox 28, the downloads are here.

For me, the colors just added insult to injury. If they had left the icons alone, I'd probably have kept the upgrade. Locking the back and reload icons to where they put them in the address bar means learning a whole new layout from where I've had them for at least 10 years.

I'm OK with the latest Firefox upgrade (version 29.0). I just need the URL field and the search field to be up there somewhere, and they are.

Something to do is run the ESR version of FF. The layout stays the same and you just get security updates. Sadly as long as Apple has a significant market share your going to keep seeing copies of their bad UI design ad companies think thats what makes them sell.

You might want to try Classic Them Restorer. Have not myself. Those who might want to upgrade to Firefox 29 for security fixes but retain the look and feel of 28 and below, may be interested.

Again, I like the new Firefox and the simplicity of it. I did a couple of css tweaks to re-size the address field and still trying to find a way to center that field in the browser window.

+1. I found the old interface clunky and awkward. I think this change makes Firefox a real contender. And it may only be my impression, but the new release seems a lot snappier, at least in Linux.

I still use Chrome (Windows) or Chromium (Linux) the most, but it's tiresome to have to turn off all the "Peeping Tom" features every time they push an upgrade.

As long as i can move the buttons back left I'll be happy, having them spread out is a pain trying to slide back and forth across thr whole display trying to find the right one.

What's the ESR version?

Extended support release
http://www.mozilla.org/en-US/firefox/organizations/

Designed for organizations who don't want to roll out a whole new version every other month when they rearrange the buttons and call it a new release.

This time it's Ebay. G-d damned it, this [stuff] is getting old!
http://money.msn.com/business-news/article.aspx?feed=OBR&date=20140521&id=17638679

Once more unto the breach, dear friends, once more
Or close the wall up with our English dead

I haven't used ebay in years but I do still have account there. I'll have to reset the password just to get in.
We almost need a regulation that requires companies to delete inactive accounts after a certain amount of inactivity.

I think Photobucket and Hotmail do that, but I don't know of any other companies.

hotmail does not as I haven't used it in years but was able to go back now and login.