class OpenSSL::Cipher

Provides symmetric algorithms for encryption and decryption. The algorithms that are available depend on the particular version of OpenSSL that is installed.

Listing all supported algorithms

A list of supported algorithms can be obtained by

puts OpenSSL::Cipher.ciphers

Instantiating a Cipher

There are several ways to create a Cipher instance. Generally, a Cipher algorithm is categorized by its name, the key length in bits and the cipher mode to be used. The most generic way to create a Cipher is the following

cipher = OpenSSL::Cipher.new('<name>-<key length>-<mode>')

That is, a string consisting of the hyphenated concatenation of the individual components name, key length and mode. Either all uppercase or all lowercase strings may be used, for example:

cipher = OpenSSL::Cipher.new('aes-128-cbc')

Choosing either encryption or decryption mode

Encryption and decryption are often very similar operations for symmetric algorithms, this is reflected by not having to choose different classes for either operation, both can be done using the same class. Still, after obtaining a Cipher instance, we need to tell the instance what it is that we intend to do with it, so we need to call either

cipher.encrypt

or

cipher.decrypt

on the Cipher instance. This should be the first call after creating the instance, otherwise configuration that has already been set could get lost in the process.

Choosing a key

Symmetric encryption requires a key that is the same for the encrypting and for the decrypting party and after initial key establishment should be kept as private information. There are a lot of ways to create insecure keys, the most notable is to simply take a password as the key without processing the password further. A simple and secure way to create a key for a particular Cipher is

cipher = OpenSSL::Cipher.new('aes-256-cfb')
cipher.encrypt
key = cipher.random_key # also sets the generated key on the Cipher

If you absolutely need to use passwords as encryption keys, you should use Password-Based Key Derivation Function 2 (PBKDF2) by generating the key with the help of the functionality provided by OpenSSL::PKCS5.pbkdf2_hmac_sha1 or OpenSSL::PKCS5.pbkdf2_hmac.

Although there is Cipher#pkcs5_keyivgen, its use is deprecated and it should only be used in legacy applications because it does not use the newer PKCS#5 v2 algorithms.

Choosing an IV

The cipher modes CBC, CFB, OFB and CTR all need an “initialization vector”, or short, IV. ECB mode is the only mode that does not require an IV, but there is almost no legitimate use case for this mode because of the fact that it does not sufficiently hide plaintext patterns. Therefore

You should never use ECB mode unless you are absolutely sure that you absolutely need it

Because of this, you will end up with a mode that explicitly requires an IV in any case. Although the IV can be seen as public information, i.e. it may be transmitted in public once generated, it should still stay unpredictable to prevent certain kinds of attacks. Therefore, ideally

Always create a secure random IV for every encryption of your Cipher

A new, random IV should be created for every encryption of data. Think of the IV as a nonce (number used once) - it’s public but random and unpredictable. A secure random IV can be created as follows

cipher = ...
cipher.encrypt
key = cipher.random_key
iv = cipher.random_iv # also sets the generated IV on the Cipher

Although the key is generally a random value, too, it is a bad choice as an IV. There are elaborate ways how an attacker can take advantage of such an IV. As a general rule of thumb, exposing the key directly or indirectly should be avoided at all cost and exceptions only be made with good reason.

Calling Cipher#final

ECB (which should not be used) and CBC are both block-based modes. This means that unlike for the other streaming-based modes, they operate on fixed-size blocks of data, and therefore they require a “finalization” step to produce or correctly decrypt the last block of data by appropriately handling some form of padding. Therefore it is essential to add the output of OpenSSL::Cipher#final to your encryption/decryption buffer or you will end up with decryption errors or truncated data.

Although this is not really necessary for streaming-mode ciphers, it is still recommended to apply the same pattern of adding the output of Cipher#final there as well - it also enables you to switch between modes more easily in the future.

Encrypting and decrypting some data

data = "Very, very confidential data"

cipher = OpenSSL::Cipher.new('aes-128-cbc')
cipher.encrypt
key = cipher.random_key
iv = cipher.random_iv

encrypted = cipher.update(data) + cipher.final
...
decipher = OpenSSL::Cipher.new('aes-128-cbc')
decipher.decrypt
decipher.key = key
decipher.iv = iv

plain = decipher.update(encrypted) + decipher.final

puts data == plain #=> true

Authenticated Encryption and Associated Data (AEAD)

If the OpenSSL version used supports it, an Authenticated Encryption mode (such as GCM or CCM) should always be preferred over any unauthenticated mode. Currently, OpenSSL supports AE only in combination with Associated Data (AEAD) where additional associated data is included in the encryption process to compute a tag at the end of the encryption. This tag will also be used in the decryption process and by verifying its validity, the authenticity of a given ciphertext is established.

This is superior to unauthenticated modes in that it allows to detect if somebody effectively changed the ciphertext after it had been encrypted. This prevents malicious modifications of the ciphertext that could otherwise be exploited to modify ciphertexts in ways beneficial to potential attackers.

Associated data, also called additional authenticated data (AAD), is optionally used where there is additional information, such as headers or some metadata, that must be also authenticated but not necessarily need to be encrypted.

An example using the GCM (Galois/Counter Mode). You have 16 bytes key, 12 bytes (96 bits) nonce and the associated data auth_data. Be sure not to reuse the key and nonce pair. Reusing an nonce ruins the security guarantees of GCM mode.

key = OpenSSL::Random.random_bytes(16)
nonce = OpenSSL::Random.random_bytes(12)
auth_data = "authenticated but unencrypted data"
data = "encrypted data"

cipher = OpenSSL::Cipher.new('aes-128-gcm').encrypt
cipher.key = key
cipher.iv = nonce
cipher.auth_data = auth_data

encrypted = cipher.update(data) + cipher.final
tag = cipher.auth_tag(16)

Now you are the receiver. You know the key and have received nonce, auth_data, encrypted and tag through an untrusted network. Note that GCM accepts an arbitrary length tag between 1 and 16 bytes. You may additionally need to check that the received tag has the correct length, or you allow attackers to forge a valid single byte tag for the tampered ciphertext with a probability of 1/256.

raise "tag is truncated!" unless tag.bytesize == 16
decipher = OpenSSL::Cipher.new('aes-128-gcm').decrypt
decipher.key = key
decipher.iv = nonce
decipher.auth_tag = tag # could be called at any time before #final
decipher.auth_data = auth_data

decrypted = decipher.update(encrypted) + decipher.final

puts data == decrypted #=> true

Note that other AEAD ciphers may require additional steps, such as setting the expected tag length (auth_tag_len=) or the total data length (#ccm_data_len=) in advance. Make sure to read the relevant man page for details.