[IMC-Tech] Help with Cacert multiple domains config

mat(a)riseup.net mat at riseup.net
Fri Jan 6 08:00:37 PST 2006


So, wich line do you recomend to include in http.conf ? 
Example ?

Mat

> -----Mensaje original-----
> De: imc-tech-bounces at lists.indymedia.org 
> [mailto:imc-tech-bounces at lists.indymedia.org] En nombre de 
> xmux at riseup.net
> Enviado el: Jueves, 05 de Enero de 2006 17:16
> Para: imc-tech at lists.indymedia.org
> Asunto: Re: [IMC-Tech] Help with Cacert multiple domains config
> 
> WARNING:  LONG....BORING.....
> 
> On Wed, Jan 04, 2006 at 07:10:56PM +0000, Chris wrote:
> > 
> > I have added an example to the wiki:
> > 
> >   
> > 
> https://docs.indymedia.org/view/Sysadmin/CaCertSsl#httpd_conf_example
> > 
> > Shout if it needs more comments etc...
> > 
> > The SSLCipherSuite line could perhaps be better, I have 
> copied it from 
> > another server, not sure what I got it from originally...
> 
> The SSLCipherSuite line looks like nonsense to me.  I'll 
> break it down for people that may not be familiar with the 
> format and comment on each part.
> 
> ALL:!ADH:!EXPORT56:!RC4+RSA:+HIGH:!MEDIUM:!LOW:!SSLv2:+EXP:+eNULL
> 
> The modifier characters work like this:
> 
> 'CIPHERS'  
> 
>   With no character, the set of ciphers specified by CIPHERS is
>   explicitly added
> 
> '!CIPHERS' 
> 
>   Permanently remove the set of ciphers.  They cannot later be
>   explicitly added again either as a group or individially.
> 
> '-CIPHERS'
> 
>   Remove the set of ciphers.  Like above but you can add them
>   again by specifying them explicitly.
> 
> '+CIPHERS' 
> 
>   Move the set of ciphers to the end of the list.  Does not
>   mean 'add to list'.  That's what 'CIPHERS' without any
>   modifier character is for.
> 
> Full details in ciphers(1) manpage
> 
> So.......
> 
> 
> ALL  
> 
>   Begin with the complete set of ciphers (This is the usual 
> way to start out)
> 
> !ADH 
> 
>   Remove all ciphersuites that use anonymous diffie-hellman 
> key exchange.
>   This is a good idea because this key exchange is not 
> authenticated with
>   certificates and can not protect against active attacks. 
> 
> !EXPORT56
> 
>   This removes all symmetric ciphers (the ones that are 
> actually used to
>   encrypt data) that have been weakened to use 56 bit keys.  
> This doesn't
>   include DES which normally uses a 56 bit key (use LOW 
> keyword for that)
>   or the much weaker 40 bit export weakened ciphers (EXPORT40 
> to remove
>   those).
> 
> !RC4+RSA
> 
>   Remove the RC4 symmetric algorithm when used with RSA key exchange.
>   Why?  Who knows.
> 
> +HIGH
> 
>   Move the symmetric ciphers with the strongest (more that 
> 128 bits) keys
>   to the end of the list!!?!
> 
> !MEDIUM
> 
>   Remove symmetric ciphers with 128 bit keys.  I don't know 
> why you would
>   want to do this.  128 bit keys are pretty long and nobody 
> is ever going
>   to be able to crack them without using science fiction 
> technology (like
>   DNA or quantum computers) that can break keys of any length.
>   
> !LOW
> 
>   Remove symmetric ciphers with 64 and 56 bit keys like old school DES
>   for example, but not including the specially weakened 
> export ciphers.
> 
> !SSLv2
> 
>   Disable SSL version 2.0.  Good idea, SSL 2.0 has many cryptographic
>   weaknesses the worst of which is that it can easily be 
> tricked into using
>   weak ciphers by an active attacker.
> 
> +EXP
> 
>   Move the export weakened ciphers to the end of the list.  
> Good thing this
>   is here or else the strongest ciphers would still be at the 
> end of the
>   list and the server would be configured to prefer to use 
> ciphers with
>   easily crackable 40 bit keys.  Happy accident it seems.
> 
> +eNULL
> 
>   This doesn't do anything because the NULL ciphersuite 
> (which doesn't use
>   any encyption at all!) is not included in the set of 'ALL' ciphers
>   because it is mostly just for testing.  If it had been 
> included in ALL
>   then the server would now be configured to use no encryption at all
>   whenever possible!  That's because all of the other ciphers 
> have either
>   been removed or moved to the end of the list.
> 
> 
> So here's what we're left with:
> 
> $ openssl ciphers -v 
> 'ALL:!ADH:!EXPORT56:!RC4+RSA:+HIGH:!MEDIUM:!LOW:!SSLv2:+EXP:+eNULL'
> DHE-RSA-AES256-SHA      SSLv3 Kx=DH       Au=RSA  
> Enc=AES(256)  Mac=SHA1
> DHE-DSS-AES256-SHA      SSLv3 Kx=DH       Au=DSS  
> Enc=AES(256)  Mac=SHA1
> AES256-SHA              SSLv3 Kx=RSA      Au=RSA  
> Enc=AES(256)  Mac=SHA1
> EDH-RSA-DES-CBC3-SHA    SSLv3 Kx=DH       Au=RSA  
> Enc=3DES(168) Mac=SHA1
> EDH-DSS-DES-CBC3-SHA    SSLv3 Kx=DH       Au=DSS  
> Enc=3DES(168) Mac=SHA1
> DES-CBC3-SHA            SSLv3 Kx=RSA      Au=RSA  
> Enc=3DES(168) Mac=SHA1
> EXP-EDH-RSA-DES-CBC-SHA SSLv3 Kx=DH(512)  Au=RSA  Enc=DES(40) 
>   Mac=SHA1 export
> EXP-EDH-DSS-DES-CBC-SHA SSLv3 Kx=DH(512)  Au=DSS  Enc=DES(40) 
>   Mac=SHA1 export
> EXP-DES-CBC-SHA         SSLv3 Kx=RSA(512) Au=RSA  Enc=DES(40) 
>   Mac=SHA1 export
> EXP-RC2-CBC-MD5         SSLv3 Kx=RSA(512) Au=RSA  Enc=RC2(40) 
>   Mac=MD5  export
> 
> 
> This is the list of cipher sets that the server will support 
> and the order in
> which it prefers to use them.
> 
> >From this list, there are three sets of symmetric ciphers 
> left that we support:
> 
> AES, 3DES, and export weakened ciphers with 40 bit keys.
> 
> AES may not be well supported because it is only specified 
> for TLS 1.0 (not
> SSL 3.0 or 2.0) and was not even in the TLS standard at all.  
> It wasn't added
> until 2002 (rfc 3268).
> 
> 3DES is supported by everything but it is really slow and is 
> even weaker than
> the 128 bit ciphers that you have already excluded.  Although 
> it has a 168 bit
> key, it is brute forceable as if it had a 112 bit key 
> (meet-in-the-middle attack).
> This is all academic though because nobody can brute force 
> 112 bit keys either.
> 
> Then there are the export ciphers.  At least you can talk to 
> export weakened clients
> which is the whole purpose of the export ciphers in the first 
> place.  I don't know
> if including these is a good idea or not but at least they 
> should be at the end of
> the list (which they are).
> 
> I wish I could give you some advice about what would be a 
> better list of ciphersuites
> to support but I'm right in the middle of some research to 
> answer that exact question.
> 
> Stay tuned....
> 
> 
> -xmux
> 
> _______________________________________________
> imc-tech mailing list
> imc-tech at lists.indymedia.org
> http://lists.indymedia.org/mailman/listinfo/imc-tech
> 




More information about the imc-tech mailing list