Strong SSL Security on Apache2
Published: 14-06-2015 | Last update: 27-04-2019 | Author: Remy van Elst | Text only version of this article
Table of Contents
This tutorial shows you how to set up strong SSL security on the Apache2webserver. We do this by updating OpenSSL to the latest version to mitigateattacks like Heartbleed, disabling SSL Compression and EXPORT ciphers tomitigate attacks like FREAK, CRIME and LogJAM, disabling SSLv3 and below becauseof vulnerabilities in the protocol and we will set up a strong ciphersuite thatenables Forward Secrecy when possible. We also enable HSTS and HPKP. This way wehave a strong and future proof ssl configuration and we get an A+ on the QuallyLabs SSL Test.
I've created a website with Copy-pastable strong cipherssuites for NGINX,Apache, Lighttpd and other software: https://cipherli.st. Handy if you don'twant to read this entire tutorial. This tutorial and https://cipherli.stare updated continuously as new vulnerabilities are discovered.
I've written an Open Source SSL server test. You can use it to test yourconfiguration, as an addition to the other SSL tests our there. It is opensource so you can host it yourself internally to test local resources. It isfast, shows you all the information so you can make your own informed decision(no ratings), and the results are saved so you can compare different settings.You can test your site via https://ssldecoder.org.
I've also written a handy tool which notifies you when your certificates areabout to expire. It is open source so you can host it yourself internally andthere is a hosted version available at https://certificatemonitor.org.
This tutorial works with the strict requirements of the SSL Labs test
You can find more info on the topics by following the links below:
Make sure you back up the files before editing them!
All protocols other than TLS 1.2 and TLS 1.3 are considered unsafe.
Edit the config file:
# Requires Apache 2.4.36 & OpenSSL 1.1.1SSLProtocol -all +TLSv1.3 +TLSv1.2SSLOpenSSLConfCmd Curves X25519:secp521r1:secp384r1:prime256v1# Older versions# SSLProtocol All -SSLv2 -SSLv3 -TLSv1 -TLSv1.1
All is a shortcut for
+SSLv2 +SSLv3 +TLSv1 or - when using OpenSSL 1.0.1 andlater -
+SSLv2 +SSLv3 +TLSv1 +TLSv1.1 +TLSv1.2, respectively. The above lineenables everything except SSLv2 and SSLv3. More info on the apache website
The Cipher Suite
(Perfect) Forward Secrecy ensures the integrity of a session key in the eventthat a long-term key is compromised. PFS accomplishes this by enforcing thederivation of a new key for each and every session.
This means that when the private key gets compromised it cannot be used todecrypt recorded SSL traffic.
The cipher suites that provide Perfect Forward Secrecy are those that use anephemeral form of the Diffie-Hellman key exchange. Their disadvantage is theiroverhead, which can be improved by using the elliptic curve variants.
The following two ciphersuites are recommended by me, and the latter by theMozilla Foundation.
The recommended cipher suite:
The recommended cipher suite for backwards compatibility (IE6/WinXP):
If your version of OpenSSL is old, unavailable ciphers will be discardedautomatically. Always use the full ciphersuite above and let OpenSSL pick theones it supports.
The ordering of a ciphersuite is very important because it decides whichalgorithms are going to be selected in priority. The recommendation aboveprioritizes algorithms that provide perfect forward secrecy.
Older versions of OpenSSL may not return the full list of algorithms. AES-GCMand some ECDHE are fairly recent, and not present on most versions of OpenSSLshipped with Ubuntu or RHEL.
- ECDHE+AESGCM ciphers are selected first. These are TLS 1.2 ciphers. No known attack currently target these ciphers.
- PFS ciphersuites are preferred, with ECDHE first, then DHE.
- AES 128 is preferred to AES 256. There has been discussions on whether AES256 extra security was worth the cost , and the result is far from obvious. At the moment, AES128 is preferred, because it provides good security, is really fa st, and seems to be more resistant to timing attacks.
- In the backward compatible ciphersuite, AES is preferred to 3DES. BEAST attacks on AES are mitigated in TLS 1.1 and a bove, and difficult to achieve in TLS 1.0. In the non-backward compatible ciphersuite, 3DES is not present.
- RC4 is removed entirely. 3DES is used for backward compatibility. See discussion in #RC4_weaknesses
- aNULL contains non-authenticated Diffie-Hellman key exchanges, that are subject to Man-In-The-Middle (MITM) attacks
- eNULL contains null-encryption ciphers (cleartext)
- EXPORT are legacy weak ciphers that were marked as exportable by US law
- RC4 contains ciphers that use the deprecated ARCFOUR algorithm
- DES contains ciphers that use the deprecated Data Encryption Standard
- SSLv2 contains all ciphers that were defined in the old version of the SSL standard, now deprecated
- MD5 contains all the ciphers that use the deprecated message digest 5 as the hashing algorithm
With Apache 2.2.x you have only DHE suites to work with, but they are notenough. Internet Explorer (in all versions) does not support the required DHEsuites to achieve Forward Secrecy. (Unless youre using DSA keys, but no onedoes; that's a long story.) Apache does not support configurable DH parametersin any version, but there are patches you could use if you can install fromsource.
Even if openssl can provide ECDHE the apache 2.2 in debian stable does notsupport this mechanism. You need apache 2.4 to fully support forward secrecy.
A workaround could be the usage of nginx as a reverse proxy because it fullysupports ECDHE.
Make sure you also add this line:
When choosing a cipher during an SSLv3 or TLSv1 handshake, normally the client'spreference is used. If this directive is enabled, the server's preference willbe used instead.
Forward Secrecy & Diffie Hellman Ephemeral Parameters
The concept of forward secrecy is simple: client and server negotiate a key thatnever hits the wire, and is destroyed at the end of the session. The RSA privatefrom the server is used to sign a Diffie- Hellman key exchange between theclient and the server. The pre-master key obtained from the Diffie-Hellmanhandshake is then used for encryption. Since the pre-master key is specific to aconnection between a client and a server, and used only for a limited amount oftime, it is called Ephemeral.
With Forward Secrecy, if an attacker gets a hold of the server's private key, itwill not be able to decrypt past communications. The private key is only used tosign the DH handshake, which does not reveal the pre- master key. Diffie-Hellmanensures that the pre-master keys never leave the client and the server, andcannot be intercepted by a MITM.
Apache prior to version 2.4.7 and all versions of Nginx as of 1.4.4 rely onOpenSSL for input parameters to Diffie-Hellman (DH). Unfortunately, this meansthat Ephemeral Diffie-Hellman (DHE) will use OpenSSL's defaults, which include a1024-bit key for the key-exchange. Since we're using a 2048-bit certificate, DHEclients will use a weaker key-exchange than non-ephemeral DH clients.
For Apache, there is no fix except to upgrade to 2.4.7 or later. With thatversion, Apache automatically selects a stronger key.
If you have Apache 2.4.8 or later and OpenSSL 1.0.2 or later, you can generateand specify your DH params file:
#Generate the parametersopenssl dhparam -out /etc/ssl/certsdhparam.pem 4096# Add the following to your Apache config.SSLOpenSSLConfCmd DHParameters "/etc/ssl/certs/dhparam.pem"
If you are using Apache with LibreSSL, or Apache 2.4.7 and OpenSSL 0.9.8a orlater, you can append the DHparams you generated earlier to the end of yourcertificate file. The documentation for that is here and below:
Custom DH parameters and an EC curve name for ephemeral keys, can also be addedto end of the first file configured using
SSLCertificateFile. This issupported in version 2.4.7 or later. Such parameters can be generated using thecommands
openssl dhparam and
openssl ecparam. The parameters can be addedas-is to the end of the first certificate file. Only the first file can be usedfor custom parameters, as they are applied independently of the authenticationalgorithm type.
Around May, Debian backported ECDH ciphers to work with apache 2.2, and it'spossible to get PFS: http://metadata.ftp-master.debian.org/changelogs//main/a/apache2/apache2 2.2.22-13+deb7u3changelog
> apache2 (2.2.22-13+deb7u2) wheezy; urgency=medium * Backport support for SSL ECC keys and ECDH ciphers.
HTTP Strict Transport Security
When possible, you should enable HTTP Strict Transport Security (HSTS),which instructs browsers to communicate with your site only over HTTPS.
HTTP Public Key Pinning Extension
You should also enable the HTTP Public Key Pinning Extension.
Public Key Pinning means that a certificate chain must include a whitelistedpublic key. It ensures only whitelisted Certificate Authorities (CA) can signcertificates for
*.example.com, and not any CA in your browser store.
I've written an article about it that has background theory and configurationexamples for Apache, Lighttpd and NGINX: https://raymii.org/s/articles/HTTPPublic Key Pinning Extension_HPKP.html
When connecting to a server, clients should verify the validity of the servercertificate using either a Certificate Revocation List (CRL), or an OnlineCertificate Status Protocol (OCSP) record. The problem with CRL is that thelists have grown huge and takes forever to download.
OCSP is much more lightweight, as only one record is retrieved at a time. Butthe side effect is that OCSP requests must be made to a 3rd party OCSP responderwhen connecting to a server, which adds latency and potential failures. In fact,the OCSP responders operated by CAs are often so unreliable that browser willfail silently if no response is received in a timely manner. This reducessecurity, by allowing an attacker to DoS an OCSP responder to disable thevalidation.
The solution is to allow the server to send its cached OCSP record during theTLS handshake, therefore bypassing the OCSP responder. This mechanism saves aroundtrip between the client and the OCSP responder, and is called OCSPStapling.
The server will send a cached OCSP response only if the client requests it, byannouncing support for the status_request TLS extension in its CLIENT HELLO.
Most servers will cache OCSP response for up to 48 hours. At regular intervals,the server will connect to the OCSP responder of the CA to retrieve a fresh OCSPrecord. The location of the OCSP responder is taken from the AuthorityInformation Access field of the signed certificate.
Poodle and TLS-FALLBACK-SCSV
SSLv3 allows exploiting of the POODLE bug. This is one more major reasonto disable this.
Google have proposed an extension to SSL/TLS named TLS FALLBACK SCSVthat seeks to prevent forced SSL downgrades. This is automatically enabled ifyou upgrade OpenSSL to the following versions:
- OpenSSL 1.0.1 has TLS FALLBACK SCSV in 1.0.1j and higher.
- OpenSSL 1.0.0 has TLS FALLBACK SCSV in 1.0.0o and higher.
- OpenSSL 0.9.8 has TLS FALLBACK SCSV in 0.9.8zc and higher.
The BEAST attack and RC4
In short, by tampering with an encryption algorithm's CBC - cipher blockchaining - mode's, portions of the encrypted traffic can be secretly decrypted.More info on the above link.
Recent browser versions have enabled client side mitigation for the beastattack. The recommendation was to disable all TLS 1.0 ciphers and only offerRC4. However, [RC4 has a growing list of attacks againstit],(http://www.isg.rhul.ac.uk/tls/) many of which have crossed the line fromtheoretical to practical. Moreover, there is reason to believe that the NSA hasbroken RC4, their so-called "big breakthrough."
Disabling RC4 has several ramifications. One, users with shitty browsers such asInternet Explorer on Windows XP will use 3DES in lieu. Triple- DES is moresecure than RC4, but it is significantly more expensive. Your server will paythe cost for these users. Two, RC4 mitigates BEAST. Thus, disabling RC4 makesTLS 1.0 users susceptible to that attack, by moving them to AES-CBC (the usualserver-side BEAST "fix" is to prioritize RC4 above all else). I am confidentthat the flaws in RC4 significantly outweigh the risks from BEAST. Indeed, withclient-side mitigation (which Chrome and Firefox both provide), BEAST is anonissue. But the risk from RC4 only grows: More cryptanalysis will surface overtime.
Factoring RSA-EXPORT Keys (FREAK)
FREAK is a man-in-the-middle (MITM) vulnerability discovered by a group ofcryptographers at INRIA, Microsoft Research and IMDEA. FREAK stands for"Factoring RSA-EXPORT Keys."
The vulnerability dates back to the 1990s, when the US government banned sellingcrypto software overseas, unless it used export cipher suites which involvedencryption keys no longer than 512-bits.
It turns out that some modern TLS clients - including Apple's SecureTransportand OpenSSL - have a bug in them. This bug causes them to accept RSA export-grade keys even when the client didn't ask for export-grade RSA. The impact ofthis bug can be quite nasty: it admits a 'man in the middle' attack whereby anactive attacker can force down the quality of a connection, provided that theclient is vulnerable and the server supports export RSA.
There are two parts of the attack as the server must also accept "export gradeRSA."
The MITM attack works as follows:
- In the client's Hello message, it asks for a standard 'RSA' ciphersuite.
- The MITM attacker changes this message to ask for 'export RSA'.
- The server responds with a 512-bit export RSA key, signed with its long-term key.
- The client accepts this weak key due to the OpenSSL/SecureTransport bug.
- The attacker factors the RSA modulus to recover the corresponding RSA decryption key.
- When the client encrypts the 'pre-master secret' to the server, the attacker can now decrypt it to recover the TLS 'm aster secret'.
- From here on out, the attacker sees plaintext and can inject anything it wants.
The ciphersuite offered here on this page does not enable EXPORT grade ciphers.Make sure your OpenSSL is updated to the latest available version and urge yourclients to also use upgraded software.
Logjam (DH EXPORT)
Researchers from several universities and institutions conducted a studythat found an issue in the TLS protocol. In a report the researchers report twoattack methods.
Diffie-Hellman key exchange allows that depend on TLS to agree on a shared keyand negotiate a secure session over a plain text connection.
With the first attack, a man-in-the-middle can downgrade a vulnerable TLSconnection to 512-bit export-grade cryptography which would allow the attackerto read and change the data. The second threat is that many servers and use thesame prime numbers for Diffie-Hellman key exchange instead of generating theirown unique DH parameters.
The team estimates that an academic team can break 768-bit primes and that anation-state could break a 1024-bit prime. By breaking one 1024-bit prime, onecould eavesdrop on 18 percent of the top one million HTTPS domains. Breaking asecond prime would open up 66 percent of VPNs and 26 percent of SSH servers.
Later on in this guide we generate our own unique DH parameters and we use aciphersuite that does not enable EXPORT grade ciphers. Make sure your OpenSSL isupdated to the latest available version and urge your clients to also useupgraded software. Updated browsers refuse DH parameters lower than 768/1024 bitas a fix to this.
Cloudflare has a detailed guide on logjam.
Heartbleed is a security bug disclosed in April 2014 in the OpenSSL cryptographylibrary, which is a widely used implementation of the Transport Layer Security(TLS) protocol. Heartbleed may be exploited regardless of whether the partyusing a vulnerable OpenSSL instance for TLS is a server or a client. It resultsfrom improper input validation (due to a missing bounds check) in theimplementation of the DTLS heartbeat extension (RFC6520), thus the bug's namederives from "heartbeat". The vulnerability is classified as a buffer over-read,a situation where more data can be read than should be allowed.
What versions of the OpenSSL are affected by Heartbleed?
Status of different versions:
- OpenSSL 1.0.1 through 1.0.1f (inclusive) are vulnerable
- OpenSSL 1.0.1g is NOT vulnerable
- OpenSSL 1.0.0 branch is NOT vulnerable
- OpenSSL 0.9.8 branch is NOT vulnerable
The bug was introduced to OpenSSL in December 2011 and has been out in the wildsince OpenSSL release 1.0.1 on 14th of March 2012. OpenSSL 1.0.1g released on7th of April 2014 fixes the bug.
By updating OpenSSL you are not vulnerable to this bug.
SSL Compression (CRIME attack)
The CRIME attack uses SSL Compression to do its magic, so we need to disablethat. On Apache 2.2.24+ we can add the following line to the SSL config file wealso edited above:
If you are using al earlier version of Apache and your distro has not backportedthis option then you need to recompile OpenSSL without ZLIB support. This willdisable the use of OpenSSL using the DEFLATE compression method. If you do thisthen you can still use regular HTML DEFLATE compression.
If you have applied the above config lines you need to restart apache:
# Check the config first:apache2ctl -t# Then restart:/etc/init.d/apache2 restart# If you are on RHEL/CentOS:apachectl -t/etc/init.d/httpd restart
Now use the SSL Labs test to see if you get a nice A+. And, of course,have a safe, strong and future proof SSL configuration!Tags: apache, pfs, security, ssl, ssl-labs, tls, tutorials