I am working on a php login system and I was wondering how it is possible to make a secure password entry system that asks for say the 1st, 2nd and 8th characters of your password, like many online banking systems do. How could one make this and have the passwords stored as double-salted hashes?
An idea that comes into my mind is to store every character of the password hashed on separate field or serialized:
PSEUDOCODE:
$password is the user password, $secret_word is the word you use to check single characters
function get_hashed_characters($password, $secret_word) {
$char_store = ""
for every character $char in $secret_word
$hashed_char = some_hash_function($char + $password)
$hash_store = $char_store + $hashed_char
return $hash_store
}
function check_hashed_char($password, $hash_store, $char_index, $char) {
if len($hash_store) < $char_index * $HASH_LEN + $HASH_LEN return false
$hashed_char = substr($hash_store, $char_index * $HASH_LEN, $HASH_LEN)
return true if $hashed_char is equals to some_hash_function($char + $password), false otherwise
}
UPDATE: as C4ud3x pointed out, I hash both the character I want to store concatenated with the password
This is a completely different solution sometimes used on Linux systems: Challenge-Response authentication
HSBC use this style of password system on their website:
and in their app: )*
After contacting them they said:
Ensuring the security of our systems is and will continue to be our number one priority.
All the details that are sent to and from our systems are encrypted
using high encryption levels. As long as you keep your log on
information secret, we can assure you that the service is secure. As
you will appreciate, we cannot provide further details about the
additional security measures used by Online Banking, as we must
protect the integrity of the system.
Though this could just be a reference to their use of SSL, I think it probably suggests they:
Encrypt password data in the database
Encrypt the passwords in the database and then encrypt the whole database for better security.
I think the best solution is that provided by ColOfAbRiX, as it doesn't require encryption (which without the technologies available to banks such as HSBC is probably not very secure).
This question already has answers here:
Secure hash and salt for PHP passwords
(14 answers)
Closed 7 years ago.
I have a VERY small site and recently I've been trying to make it more secure, I used to store my passwords in plain text.
I think Im doing it right, but as a "hobby" programmer I wanna make sure so I ask you, the professionals
When a user register I do: password_hash($their_password, PASSWORD_DEFAULT) and store that in the 'password' column in the users table.
I use PASSWORD_DEFAULT as that seems the best according to php.net.
Note that this constant is designed to change over time as new and stronger algorithms are added to PHP."
Sounds good!
And the Login part (very simple):
if (count($_POST) > 0) {
$username = trim($_POST['username']);
$password = trim($_POST['password']);
$query = $db->prepare("SELECT password FROM users WHERE username = ?");
$query->execute(array($username));
$row = $query->fetch();
if (password_verify($password, $row['password'])) {
echo "Correct password";
// create session...
} else {
// wrong password
}
Maybe I should check if the username exists first but other than that what do you think?
You appear to have perfectly understood the documentation and how to construct the code you need. Shame on you for using plaintext password even temporarily, but your decision to fix with the correct method (ie. not md5 like me a silly person (I really need to update my password saving systems...)) is awesome.
The only issue I can see is that some people might have their passwords start or end with a space. Such passwords would lose their leading/trailing spaces and indeed the user may be alarmed that they can log in with two spaces, or none! So probably best to remove those trim calls ;)
Hello :) I'm a hobbyist too, and I think I can point you in the right direction, despite not knowing much about how the magic happens.
1) user enters his password, program encrypts using a certain method, and THAT encrypted password gets saved.
2) bam. NOBODY can see what the original was - not even the user who entered it. At login, that operation is repeated and "whatever the user entered for login" password gets encrypted by the same process and then it is compared with the also encrypted saved password. They should match if they were the same unencrypted original.
Well
For extra security, something called a "salt" is sometimes "added" to the encryption process, to make it yet harder to crack the password. Say somehow someone got hold of your encryption code AND encrypted passwords list, and tries to revert the process by reverse engineering your code? Well, now that person has the extra job of finding what your "salt" was... (it could be a string saved in your server, a clever "playing with dates of the month" trick, etc... many options). This is what I remember from what I read. Plenty of pointers to get you started. And more:
I use this: which I got from somewhere in the internet years ago
function encryptTheString($password, $salt, $iter_count=4096, $keylen=64, $hash_alg= 'sha256' )
{
// Compute the length of hash alg output.
// Some folks use a static variable and save the value of the hash len.
// Considering we are doing 1000s hmacs, doing one more won't hurt.
$hashlen = strlen(hash($hash_alg, null, true));
// compute number of blocks need to make $keylen number of bytes
$numblocks = ceil($keylen / $hashlen);
// blocks are appended to this
$output = '';
for ($i = 1; $i <= $numblocks; ++$i) {
$block = hash_hmac($hash_alg, $salt . pack('N', $i), $password, true);
$ib = $block;
for ($j = 1; $j < $iter_count; ++$j) {
$block = hash_hmac($hash_alg, $block, $password, true);
$ib ^= $block;
}
$output .= $ib;
}
// extract the right number of output bytes
return substr($output, 0, $keylen);
}
And a call like
$ePassword=ANDYETpbkdf2($password,"111111111122222222223333333333444444444455555555566666666661234");
Would be perfectly ok :) give sha256 a reading for the start of further enlightenment.
I was wondering how mobile authenticators work (like Battle.net, rift, some banks have one, etc.), so I can make one for my own site (just for fun).
I understand the basics: authenticator has code related to the phone and code related to the website. Users enters the phone code on the website. Can then generate a token related (using the phone and website code).
I'm just wondering how the tokens are created. Is there a standard algorithm for this? How does the algorithm work? Any existing PHP libraries that can do something like this (as an example)?
Have a look at Google Authenticator. There are already iPhone, Android and Blackberry apps for that and it's an established protocol.
They have implemented it as an open-source PAM module which you may be able to use with the PECL PAM package.
There is a pure PHP version but I haven't used that so can't vouch for it.
The spec isn't that complex so you could probably implement it yourself, especially if you converted the C module. The specification linked there explains its working in full detail.
Edit: I guess to answer the original question, that's an RFC, so it's somewhat standardised, and it's a fully open specification and the tools to use it are fully open-source. The protocols are known as HOTP and TOTP. The former is HMAC based on a counter (so the nth password is used) whereas the latter is time-based (so the password cycles every 30 seconds).
Concerning the Blizzad Battle.Net authenticator, you can find an open source implementation in PHP : https://github.com/krtek4/php-bma
The implementation is used to provide a online authentication service for Battle.Net : https://authenticator.me
If you want to do something like it for your website, it's pretty simple. The only thing to share between the server and client part are the secret generated by the server. So when a client is requesting for a new secret, just store it and you will be able to compute the code at any moment to compare with what is sent to you.
I implemented this once. I use a 4 digit key with a subset of characters (notice that potentially confusing characters like 0oO and l1L are removed. I used 4 characters because the potential space of 4 digits from the characters set was larger than the 6 digits of an RSA key.
Anyway, I let the user log in with their username and password. If that is correct, generate a key and send it to the phone and save it in the session and show the user the next page, which requires the key be entered. The user gets the 4 digit key from their phone and enters it into the page. Then check what they entered against the session-saved key and there you go.
Some handy features to have: make the key expire after a few minutes, but long enough that text message delays don't make it impossible. Make it expire after a few bad tries. Give the users a link to resend the key or to send a new key.
//pick a random 4 digit string
$chars = "abcdefghjkrstwxyzABCDEFGHJKRSTWXYZ23456789";
$key = "";
for($i=0;$i<4;$i++){
//here, rand is used, but any generator could be used
//to choose the characters.
$key .= $chars[rand(0,strlen($chars)-1)];
}
//save it to the session
$_SESSION['test']['KEY'] = $key;
If it were me I'd go with generating a hash based on the previously used hash and a common nonce, the tricky bit would be keeping the two systems in sync. e.g.
<?php
class otp {
var $salt;
var $previous_hash;
var $user_id;
function __construct($user_id)
{
$this->user_id=$user_id;
list($this->$salt, $this->$previous_hash)
=unserialize(file_get_contents(BASE_PATH . $user_id));
}
function authenticate($submitted_otp)
{
if (!$this->salt) {
// user does not exist
return false;
}
$new_hash=$this->previous_hash;
// allow for the sequence to get out of sync by 5 steps....
for ($x=0; $x<5; $x++) {
$new_hash=md5($this->salt, $new_hash);
if ($new_hash===$submitted_otp) {
$this->update_token($new_hash);
return true;
}
}
// none of the next N iterations of the local password match
return false;
}
function create_user($user_id, $salt, $init_hash)
{
return file_put_contents(BASE_PATH . $user_id, array($salt, $init_hash));
}
function update_token($new_hash)
{
file_put_contents(BASE_PATH . $user_id, array($this->salt, $new_hash));
}
}
Of course, in practice you probably wouldn't want to use a whole 32 char md5 hash (just, say, the first 6 characters, and applying cleansing such as changing 'S' to '5' etc).
I was looking for an effective algorithm that can give me an accurate idea of how strong a password is.
I found that several different websites use several different algorithms as I get different password strength ratings on different websites.
This has grown to my general brain dump of best practices for working with passwords in PHP/MySQL.
The ideas presented here are generally not my own, but the best of what I've found to date.
Ensure you are using SSL for all operations involving user information. All pages that involve these forms should check they are being called via HTTPS, and refuse to work otherwise.
You can eliminate most attacks by simply limiting the number of failed logins allowed.
Allow for relatively weak passwords, but store the number of failed logins per user and require a captcha or password verification by email if you exceed it. I set my max failures to 5.
Presenting login failures to the user needs to be carefully thought out as to not provide information to attackers.
A failed login due to a non existent user should return the same message as a failed login due to a bad password. Providing a different message will allow attackers to determine valid user logins.
Also make sure you return exactly the same message in the event of a failure for too many logins with a valid password, and a failure with too many logins and a bad password. Providing a different message will allow attackers to determine valid user passwords. A fair number of users when forced to reset their password will simply put it back to what it was.
Unfortunately limiting the number of logins allowed per IP address is not practical. Several providers such as AOL and most companies proxy their web requests. Imposing this limit will effectively eliminate these users.
I've found checking for dictionary words before submit to be inefficient as either you have to send a dictionary to the client in javascript, or send an ajax request per field change. I did this for a while and it worked ok, but didn't like the traffic it generated.
Checking for inherently weak passwords minus dictionary words IS practical client side with some simple javascript.
After submit, I check for dictionary words, and username containing password and vice versa server side. Very good dictionaries are readily downloadable and the testing against them is simple. One gotcha here is that to test for a dictionary word, you need to send a query against the database, which again contains the password. The way I got around this was to encrypt my dictionary before hand with a simple encryption and end positioned SALT and then test for the encrypted password. Not ideal, but better than plain text and only on the wire for people on your physical machines and subnet.
Once you are happy with the password they have picked encrypt it with PHP first, then store. The following password encryption function is not my idea either, but solves a number of problems. Encrypting within PHP prevents people on a shared server from intercepting your unencrypted passwords. Adding something per user that won't change (I use email as this is the username for my sites) and add a hash (SALT is a short constant string I change per site) increases resistance to attacks. Because the SALT is located within the password, and the password can be any length, it becomes almost impossible to attack this with a rainbow table.
Alternately it also means that people can't change their email and you can't change the SALT without invalidating everyone's password though.
EDIT: I would now recommend using PhPass instead of my roll your own function here, or just forget user logins altogether and use OpenID instead.
function password_crypt($email,$toHash) {
$password = str_split($toHash,(strlen($toHash)/2)+1);
return hash('sha256', $email.$password[0].SALT.$password[1]);
}
My Jqueryish client side password meter. Target should be a div. It's width will change between 0 and 100 and background color will change based on the classes denoted in the script. Again mostly stolen from other things I've found:
$.updatePasswordMeter = function(password,username,target) {
$.updatePasswordMeter._checkRepetition = function(pLen,str) {
res = ""
for ( i=0; i<str.length ; i++ ) {
repeated=true;
for (j=0;j < pLen && (j+i+pLen) < str.length;j++)
repeated=repeated && (str.charAt(j+i)==str.charAt(j+i+pLen));
if (j<pLen) repeated=false;
if (repeated) {
i+=pLen-1;
repeated=false;
}
else {
res+=str.charAt(i);
};
};
return res;
};
var score = 0;
var r_class = 'weak-password';
//password < 4
if (password.length < 4 || password.toLowerCase()==username.toLowerCase()) {
target.width(score + '%').removeClass("weak-password okay-password good-password strong-password"
).addClass(r_class);
return true;
}
//password length
score += password.length * 4;
score += ( $.updatePasswordMeter._checkRepetition(1,password).length - password.length ) * 1;
score += ( $.updatePasswordMeter._checkRepetition(2,password).length - password.length ) * 1;
score += ( $.updatePasswordMeter._checkRepetition(3,password).length - password.length ) * 1;
score += ( $.updatePasswordMeter._checkRepetition(4,password).length - password.length ) * 1;
//password has 3 numbers
if (password.match(/(.*[0-9].*[0-9].*[0-9])/)) score += 5;
//password has 2 symbols
if (password.match(/(.*[!,#,#,$,%,^,&,*,?,_,~].*[!,#,#,$,%,^,&,*,?,_,~])/)) score += 5;
//password has Upper and Lower chars
if (password.match(/([a-z].*[A-Z])|([A-Z].*[a-z])/)) score += 10;
//password has number and chars
if (password.match(/([a-zA-Z])/) && password.match(/([0-9])/)) score += 15;
//
//password has number and symbol
if (password.match(/([!,#,#,$,%,^,&,*,?,_,~])/) && password.match(/([0-9])/)) score += 15;
//password has char and symbol
if (password.match(/([!,#,#,$,%,^,&,*,?,_,~])/) && password.match(/([a-zA-Z])/)) score += 15;
//password is just a nubers or chars
if (password.match(/^\w+$/) || password.match(/^\d+$/) ) score -= 10;
//verifing 0 < score < 100
score = score * 2;
if ( score < 0 ) score = 0;
if ( score > 100 ) score = 100;
if (score > 25 ) r_class = 'okay-password';
if (score > 50 ) r_class = 'good-password';
if (score > 75 ) r_class = 'strong-password';
target.width(score + '%').removeClass("weak-password okay-password good-password strong-password"
).addClass(r_class);
return true;
};
Fundamentally you want to prevent to major types of attacks
Dictionary attacks
Brute force attacks
To prevent the first, you want to consider passwords containing common words weak. To prevent the second, you want to encourage passwords of reasonable length (8+ characters is common) and with a reasonably large character set (include letters, numbers, and special characters). If you consider lower case and upper case letters to be different, that increases the character set substantially. However, this creates a usability issue for some user communities so you need to balance that consideration.
A quick google search turned up solutions that account for brute force attacks (complex password) but not for dictionary attacks. PHP Password Strength Meter from this list of strength checkers runs the check server-side, so it could be extended to check a dictionary.
EDIT:
By the way... you should also limit the number of login attempts per user. This will make both types of attacks less likely. Effective but not-user-friendly is to lock an account after X bad attempts and require a password reset. More user friendly but more effort is to throttle time between login attempts. You can also require CAPTCHA after the first few login attempts (which is something that Stack Overflow requires after too many edits, or for very new users).
Basically you probably want to use Regular Expressions to validate the length and complexity of the password.
A good example doing this using javascript can be found here:
http://marketingtechblog.com/programming/javascript-password-strength/
As Daren Schwenke pointed it out, you'd better work on the security yourself and not put this in the user hands.
But it's good to provide some hints to the user of how strong his password is, because the best way to get a password is still social engenering.
So you can hack a little client side script that checks the user password strenght as a courtesy indicator, in real time. It blocks nothing, but gives him a good warm feeling when it turns green :-)
Basically what you must check is commom sense : check if the password contains letters, numbers and non alphabetical caracters, in a reasonable quantity.
You can hack your own algo very easily : just make 10 / 10 mark :
0 is a zero lenght password;
+2 for every 8 caracters in the password (15 is supposed to be a safe lenght);
+1 for the use of a letter, +2 for the use of 2 letters;
+1 for the use of a number, +2 for the use of 2 numbers;
+1 for the use of a non alphabetical caracters, +2 for 2.
You don't need to check for godlike passwords (are there capitalized letters, where are positioned the special caracters, etc), your users are not in the bank / military / secret service / monthy python movies industry, are they ?
You can code that in an hour in without crazy javascript skills.
And anyway, valid the password and move all the security code on the server side. If you can delegate authentification (e.g : open ID), even better.
Don't Roll-Your-Own!
Cryptography experts discourage roll-your-own cryptography for reasons that should be obvious.
For the very same reasons, one should not attempt to roll his own solution to the problem of measuring a password's strength; it is very much a cryptographic problem.
Don't get into the ugly business of authoring some massive regular expression for this purpose; you will likely fail to account for several factors that influence a password's overall strength.
It's a Difficult Problem
There is considerable difficulty inherent to the problem of measuring a password's strength. The more research I perform on this subject, the more I realize that this is a "unidirectional" problem; that is, one cannot measure the "difficulty" (computational cost) of cracking a password efficiently. Rather, it is more efficient to provide complexity requirements and measure the password's ability to meet them.
When we consider the problem logically, a "crackability index" doesn't make much sense, as convenient as it sounds. There are so many factors that drive the calculation, most of which relate to the computational resources devoted to the cracking process, so as to be impractical.
Imagine pitting John the Ripper (or a similar tool) against the password in question; it might take days to crack a decent password, months to crack a good password, and until the sun burns-out to crack an exceptional password. This is not a practical means by which to measure password strength.
Approaching the problem from the other direction is far more manageable: if we supply a set of complexity requirements, it's possible to judge the relative strength of a password very quickly. Obviously, the supplied complexity requirements must evolve over time, but there are far fewer variables for which to account if we approach the problem in this way.
A Viable Solution
There is a standalone utility available from Openwall entitled passwdqc (presumably, standing for Password Quality Checker). Openwall developer, Solar Designer, does appear to be a bona fide cryptography expert (his works speak for themselves), and so is qualified to author such a tool.
For my particular use-case, this is a far more attractive solution than using an ill-conceived JavaScript snippet living in some dark corner of the Web.
Establishing parameters for your particular needs is the hardest part. The implementation is the easy part.
A Practical Example
I offer a simple implementation in PHP to provide a jump-start. Standard disclaimers apply.
This example assumes that we're feeding an entire list of passwords to the PHP script. It goes without saying that if you are doing this with real passwords (e.g., those dumped out of a password manager), extreme caution should be exercised with regard to password-handling. Simply writing the unencrypted password dump to disk jeopardizes the security of your passwords!
passwords.csv:
"Title","Password"
"My Test Password","password123"
"Your Test Password","123456!!!"
"A strong password","NFYbCoHC5S7dngitqCD53tvQkAu3dais"
password-check.php:
<?php
//A few handy examples from other users:
//http://php.net/manual/en/function.str-getcsv.php#117692
$csv = array_map('str_getcsv', file('passwords.csv'), [',']);
array_walk($csv, function(&$a) use ($csv) {
$a = array_combine($csv[0], $a);
});
//Remove column header.
array_shift($csv);
//Define report column headers.
$results[] = [
'Title',
'Result',
'Exit Code',
];
$i = 1;
foreach ($csv as $p) {
$row['title'] = $p['Title'];
//If the value contains a space, it's considered a passphrase.
$isPassphrase = stristr($p['Password'], ' ') !== false ? true : false;
$cmd = 'echo ' . escapeshellarg($p['Password']) . ' | pwqcheck -1 min=32,24,22,20,16 max=128';
if ($isPassphrase) {
$cmd .= ' passphrase=3';
}
else {
$cmd .= ' passphrase=0';
}
$output = null;
$exitCode = null;
$stdOut = exec($cmd, $output, $exitCode);
//Exit code 0 represents an unacceptable password (not an error).
//Exit code 1 represents an acceptable password (it meets the criteria).
if ($exitCode === 0 || $exitCode === 1) {
$row['result'] = trim($stdOut);
$row['exitCode'] = $exitCode;
}
else {
$row['result'] = 'An error occurred while calling pwqcheck';
$row['exitCode'] = null;
}
$results[$i] = $row;
$i++;
}
$reportFile = 'report.csv';
$fp = #fopen($reportFile, 'w');
if ($fp !== false) {
foreach ($results as $p) {
fputcsv($fp, $p);
}
fclose($fp);
}
else {
die($reportFile . ' could not be opened for writing (destination is not writable or file is in use)');
}
exit;
Resultant report.csv:
Title,Result,"Exit Code"
"My Test Password","Bad passphrase (too short)",1
"Your Test Password","Bad passphrase (too short)",1
"A strong password",OK,0
Wrapping-Up
I have yet to find a more thorough solution on the Web; needless to say, I welcome any other recommendations.
Obviously, this approach is not ideal for certain use-cases (e.g., a "password strength meter" implemented "client-side"). Even so, it would be trivial to make an AJAX call to a server-side resource that returns a pass/fail response using the approach outlined above, but such an approach should assume the potential for abuse (e.g., DoS attacks) and would require secure communication between client and server, as well as acceptance of the risks associated with transmitting the un-hashed password.
I can't think of a specific algorithm to check the strengh of a password. What we do is we define several criterion and when the password respect a criteria, we add 1 to its score. When the password reach a threshold, the password is strong. Otherwise it is weak.
You can define many different level of strengh if with different throeshold, or you can define different value for a specific criteria. For example, if a password has 5 character, we add 1, but if it got 10, then we add 2.
here is a list of criterion to check for
Length (8 to 12 is ok, more is better)
Contains lowercase letter
Contains uppercase letter
The upper case letter is NOT the first one.
Contains number
Contains symbols
the last character is NOT a human like symbol (ex : . or !)
Does not look like a dictionnary word. Some wise password crack contains library of word and letter substitutes (like Library --> L1br#ry )
Hope that help.
Currently my database user and its password are pretty easily to guess, eg.
database user: dbadmin
database pwd : super + companyname
What and how to generate a secure a secure database password? Using md5 or sha1??
What are the things that I need to pay attention to secure my database?
I am using php, thanks
If what you're looking for is just a secure password generator, there are a number of tools for creating strong passwords. Keepass is one that I use, but you can also look into pwgen, apg, passgen, or others.
Keepass -- http://keepass.info/
apg -- http://www.debianadmin.com/automated-password-generator-in-debian.html
passgen -- http://www.linuxbuilt.com/passwords.php
pwgen -- http://sourceforge.net/projects/pwgen/
To keep the database secure you also need to consider where you're using the username/password combination in your scripts. One technique that I've seen used often is to put the credentials in a separate file and to import them everywhere else they're needed. This way you can put strict access regulations on that one file through your webserver and filesystem security settings.
Security is a layered approach, and the more precautions you take the more resistant your server will be to malicious activity.
Use some kind of pattern to create a password that is complex, but also possible for you to remember.
E.g. Think of a silly sentence and take the first letter of every word.
"16 Butterflys drew straws to see which 5 should become Caterpillars first."
forms the password "16Bdstsw5sbCf".
That's 13 chars long, contains 3 numbers and some upper case chars. That should be pretty strong, and it's much easier to remember than just a random string.
For even better strength, throw some punctuation in there too.
If you use phpMyAdmin it has, when you are creating a user, a 'generate password' option. Try that or you could do something like this:
function generatePassword($length) {
// start with a blank password
$password = "";
// define possible characters
$possible = "0123456789bcdfghjkmnpqrstvwxyz";
// set up a counter
$i = 0;
// add random characters to $password until $length is reached
while ($i < $length) {
// pick a random character from the possible ones
$char = substr($possible, mt_rand(0, strlen($possible)-1), 1);
// we don't want this character if it's already in the password
if (!strstr($password, $char)) {
$password .= $char;
$i++;
}
}
// done!
return $password;
}
You can make a secure password simply by mashing the keyboard:
Z3w947CFqnY39cfo
That's as secure as you possibly need - it's immune to dictionary attacks, and long enough to be immune to brute force attacks.
(Of course you'll need to write it down, unless you have a trick memory, but that's always a compromise with passwords - and presumably it will appear in your application code or config anyway.)