Hashing Passwords

Websites and other systems should never store passwords in cleartext. Instead, password hashes should be stored. Password salting and algorithm choice are important.

Learning Objectives

You should be able to:

• Describe password salting
• Select a good algorithm for password storage

Video Walkthrough

Use this video to follow along with the steps in this lab.

Hashing Passwords

Like files, passwords can be hashed. When you submit your password on a website, the website should compare the hash of the submitted password to the hash of your password that the website stored when you created your account. Websites should never store passwords in cleartext.

There are 2 recommendations for hashing passwords that differ from hashing files:

1) Using salt, and 2) Selecting slower algorithms.

Salting Passwords

In the context of password hashing, salt is random data added before the password. The salt will be stored in the database with your username and password hash.

• Launch your Linux virtual machine and access the terminal.
• Run the following command to compute the SHA2-256 hash of the password "pass123."

printf "pass123" | sha256sum

• (The printf function is used here to send the quoted text directly to the sha256sum program without any carriage returns or line breaks.) The result should be:

9b8769a4a742959a2d0298c36fb70623f2dfacda8436237df08d8dfd5b37374c

• There is a problem, though. If two people in the database use the same password, the same password hash will show up several times. In fact, if that password has ever been used in another system and somebody has cracked it, everybody knows that somebody in the database with the password hash 9b8769a4a742959a2d0298c36fb70623f2dfacda8436237df08d8dfd5b37374c has the password pass123.
• Mentally roll some dice and come up with 6 random digits. Add them to the beginning of the previous password to salt it. Compute the hash.

printf "829974pass123" | sha256sum

The resulting hash of the above is:

ffc90137b42e823fc2aa04adcdfcdf19bd6d8747dac9e4d4dfd52bef24423177

• The user does not need to see the random salt. The salt could be stored in a database like so:

Username	Salt	Hash
joe123	829974	ffc90137b42e823fc2aa04adcdfcdf19bd6d8747dac9e4d4dfd52bef24423177
sue789	266714	f04af89361e83062d6dad7c812f65dc42f560dc091d704f385beb7dc6878e1a5
eve765	109872	893447cf95bd33a89834643c0265da12030daef39729d4143c83791c09168529

Each of the 3 users in the table above has the password "pass123," but because they have random salt, the password hashes are completely different. Password salting makes cracking passwords much more difficult.

Slower Algorithms

File hashing algorithms are optimized for speed. But when it comes to hashing passwords, slower can actually be better. If an attacker is trying to crack passwords, the faster the algorithm runs, the faster the passwords are cracked. Certain algorithms intentionally slow down the password hashing so that it makes it difficult for passwords to be cracked. Some of these algorithms also salt passwords by default.

• In the Linux terminal, launch Python (specifying Python version 3).

python3

• Python version 3 should be the default. The bcrypt library should be present.
• Run the following code to create a bcrypt hash of the password "pass123."

import bcrypt
bcrypt.hashpw(b"pass123", bcrypt.gensalt())

Python will calculate the hash. Your results will look similar to the following, but because of the random salt, your value will be different.

b'$2b$12$SDo.HaGB8HTqhD0WnjL/KuToKXHDL/RKBZ9bd0ikGGiX9wZwkNDF6'

• Press the up arrow key and enter to run the previous line of Python code again. And do that again. And again. See how long it takes to run.

• In Python, run exit() to quit the Python interactive interpreter.

exit()

• Notice that the bcrypt hashes were not generated instantly. It takes a fraction of a second, but a long time in computer time. Bcrypt intentionally wastes some of your computing power so that it goes slower. Depending on computer hardware, bcrypt can make about 30 hashes per second. That same hardware can do thousands of SHA2-256 hashes. Scrypt and PBKDF2 (password-based key derivation function version 2) work similarly to bcrypt.

Reflection

• How do websites store passwords?
  • Search for details on data breaches by looking for MD5, SHA1, SHA256, bcrypt, salting, etc.
• Does choosing good passwords still matter if you do not know how websites store their passwords?
• Why is password reuse a problem?

Key Terms

• Password Hash: A fixed-size string or number generated from a password using a cryptographic hash function. It is used to securely store passwords by converting them into a format that is difficult to reverse-engineer. When a user logs in, the entered password is hashed and compared to the stored hash to verify the user's identity.
• Salt (Hashing): A random value added to a password before hashing to ensure that identical passwords produce different hash values. Salting helps protect against dictionary attacks and rainbow table attacks by making it more difficult for attackers to precompute hash values for common passwords.
• SHA2-256: A cryptographic hash function that produces a 256-bit (32-byte) hash value, commonly represented as a 64-character hexadecimal string. It is part of the SHA-2 (Secure Hash Algorithm 2) family and is widely used for secure hashing in various applications, including password hashing and data integrity verification.
• Bcrypt: A password hashing function designed to be computationally intensive to resist brute-force attacks. Bcrypt incorporates a salt to protect against rainbow table attacks and allows the hashing process to be adjusted to increase computational cost, making it more difficult for attackers to crack passwords using modern hardware.