Re: [CentOS] SELinux breaks Squid's ssl_crtd helper

[Nicolas Kovacs]

Le 10/03/2018 à 18:18, Gordon Messmer a écrit :
> Start by running "ausearch -c 'ssl_crtd' --raw" by itself.  Try to
> determine whether or not all of the affected files are mentioned in that
> output.
> 
> Typically, to generate a complete policy, you'll need to run in
> permissive mode while you operate the system, so that all of the things
> that you want to allow are recorded.  Many services that need a new
> policy will generate more than one AVC denial, and in enforcing mode
> they'll terminate or at least cease processing the labeled resources
> that they need after the first denial.  In permissive mode, you should
> get a better list of exceptions that are required, because AVCs are
> recorded, but the application isn't actually denied permission to those
> resources.
> 
> When your logs are complete, remove the old module and generate a new
> one according to the directions from sealert.

OK, I found the solution. This is actually a bug in Squid's default
SELinux policy, but it can be corrected manually.

https://blog.microlinux.fr/squid-https-centos/#configuration

Cheers,

Niki

-- 
Microlinux - Solutions informatiques durables
7, place de l'église - 30730 Montpezat
Site : https://www.microlinux.fr
Blog : https://blog.microlinux.fr
Mail : i...@microlinux.fr
Tél. : 04 66 63 10 32
___
CentOS mailing list
CentOS@centos.org
https://lists.centos.org/mailman/listinfo/centos

Re: [CentOS] SELinux breaks Squid's ssl_crtd helper

[Gordon Messmer]

On 03/09/2018 05:18 AM, Nicolas Kovacs wrote:

Do allow this
access for now by executing:
# ausearch -c 'ssl_crtd' --raw | audit2allow -M my-sslcrtd
# semodule -i my-sslcrtd.pp

Unfortunately the suggested solution doesn't work



Start by running "ausearch -c 'ssl_crtd' --raw" by itself.  Try to 
determine whether or not all of the affected files are mentioned in that 
output.


Typically, to generate a complete policy, you'll need to run in 
permissive mode while you operate the system, so that all of the things 
that you want to allow are recorded.  Many services that need a new 
policy will generate more than one AVC denial, and in enforcing mode 
they'll terminate or at least cease processing the labeled resources 
that they need after the first denial.  In permissive mode, you should 
get a better list of exceptions that are required, because AVCs are 
recorded, but the application isn't actually denied permission to those 
resources.


When your logs are complete, remove the old module and generate a new 
one according to the directions from sealert.



___
CentOS mailing list
CentOS@centos.org
https://lists.centos.org/mailman/listinfo/centos

Re: [CentOS] SELinux breaks Squid's ssl_crtd helper

[peter.winterflood]

The ausearch is only identifying issues that there are logs for at that time.
Best to create a file context for that path permanantly.
Theres an example of how to do this in the selinux man pages. Typing on my 
phone im tryinf to remember the command that the example in its man page.  
Semanage fcontext is what you are looking for followed by a restorecon on 
the path you added.

Regards peter

Sent with AquaMail for Android
http://www.aqua-mail.com


On 9 March 2018 13:18:45 "Nicolas Kovacs"  wrote:


Hi,

I've setup a transparent HTTP+HTTPS proxy on my server running CentOS 7,
using Squid. Here's my configuration file.

--8<
# /etc/squid/squid.conf



# Définitions

acl localnet src 192.168.2.0/24

acl SSL_ports port 443

acl Safe_ports port 80  # http

acl Safe_ports port 21  # ftp

acl Safe_ports port 443 # https

acl Safe_ports port 70  # gopher

acl Safe_ports port 210 # wais

acl Safe_ports port 1025-65535  # unregistered ports

acl Safe_ports port 280 # http-mgmt
acl Safe_ports port 488 # gss-http
acl Safe_ports port 591 # filemaker
acl Safe_ports port 777 # multiling http
acl CONNECT method CONNECT

# Règles d'accès
http_access deny !Safe_ports
http_access deny CONNECT !SSL_ports
http_access allow localhost manager
http_access deny manager
http_access allow localnet
http_access allow localhost
http_access deny all

# Ports du proxy
http_port 3130
http_port 3128 intercept
https_port 3129 intercept ssl-bump \
  cert=/etc/squid/ssl_cert/nestor.microlinux.lan.pem \
  generate-host-certificates=on dynamic_cert_mem_cache_size=4MB

# Emplacement de ssl_crtd et du cache des certificats TLS
sslcrtd_program /usr/lib64/squid/ssl_crtd -s /var/lib/ssl_db -M 4MB
sslcrtd_children 8 startup=1 idle=1

# SSL-Bump
acl step1 at_step SslBump1
ssl_bump peek step1
ssl_bump bump all

# Taille du cache dans la RAM
cache_mem 2048 MB

# Emplacement et taille du cache sur le disque
# cache_dir ufs /var/spool/squid 100 16 256

# Vidage système
coredump_dir /var/spool/squid

# Durée de vie des fichiers sans date d'expiration
refresh_pattern ^ftp:   144020% 10080
refresh_pattern ^gopher:14400%  1440
refresh_pattern -i (/cgi-bin/|\?) 0 0%  0
refresh_pattern .   0   20% 4320
--8<

First, I tested the configuration in permissive mode, where it works
perfectly. After running it a while, I get the following SELinux alert.

# sealert -a /var/log/audit/audit.log
* Plugin catchall (17.1 confidence) suggests *
If you believe that ssl_crtd should be allowed read access on the
index.txt file by default. Then you should report this as a bug. You can
generate a local policy module to allow this access. Do allow this
access for now by executing:
# ausearch -c 'ssl_crtd' --raw | audit2allow -M my-sslcrtd
# semodule -i my-sslcrtd.pp

Unfortunately the suggested solution doesn't work, because the problem
reappears, and I get all sorts of errors in /var/log/squid/cache.log,
all due to ssl_crtd not being able to access stuff under /var/lib/ssl_db.

Any idea how I can solve this problem permanently ?

Cheers,

Niki

--
Microlinux - Solutions informatiques durables
7, place de l'église - 30730 Montpezat
Site : https://www.microlinux.fr
Blog : https://blog.microlinux.fr
Mail : i...@microlinux.fr
Tél. : 04 66 63 10 32
___
CentOS mailing list
CentOS@centos.org
https://lists.centos.org/mailman/listinfo/centos



___
CentOS mailing list
CentOS@centos.org
https://lists.centos.org/mailman/listinfo/centos

[CentOS] SELinux breaks Squid's ssl_crtd helper

[Nicolas Kovacs]

Hi,

I've setup a transparent HTTP+HTTPS proxy on my server running CentOS 7,
using Squid. Here's my configuration file.

--8<
# /etc/squid/squid.conf



# Définitions

acl localnet src 192.168.2.0/24

acl SSL_ports port 443

acl Safe_ports port 80  # http

acl Safe_ports port 21  # ftp

acl Safe_ports port 443 # https

acl Safe_ports port 70  # gopher

acl Safe_ports port 210 # wais

acl Safe_ports port 1025-65535  # unregistered ports

acl Safe_ports port 280 # http-mgmt
acl Safe_ports port 488 # gss-http
acl Safe_ports port 591 # filemaker
acl Safe_ports port 777 # multiling http
acl CONNECT method CONNECT

# Règles d'accès
http_access deny !Safe_ports
http_access deny CONNECT !SSL_ports
http_access allow localhost manager
http_access deny manager
http_access allow localnet
http_access allow localhost
http_access deny all

# Ports du proxy
http_port 3130
http_port 3128 intercept
https_port 3129 intercept ssl-bump \
  cert=/etc/squid/ssl_cert/nestor.microlinux.lan.pem \
  generate-host-certificates=on dynamic_cert_mem_cache_size=4MB

# Emplacement de ssl_crtd et du cache des certificats TLS
sslcrtd_program /usr/lib64/squid/ssl_crtd -s /var/lib/ssl_db -M 4MB
sslcrtd_children 8 startup=1 idle=1

# SSL-Bump
acl step1 at_step SslBump1
ssl_bump peek step1
ssl_bump bump all

# Taille du cache dans la RAM
cache_mem 2048 MB

# Emplacement et taille du cache sur le disque
# cache_dir ufs /var/spool/squid 100 16 256

# Vidage système
coredump_dir /var/spool/squid

# Durée de vie des fichiers sans date d'expiration
refresh_pattern ^ftp:   144020% 10080
refresh_pattern ^gopher:14400%  1440
refresh_pattern -i (/cgi-bin/|\?) 0 0%  0
refresh_pattern .   0   20% 4320
--8<

First, I tested the configuration in permissive mode, where it works
perfectly. After running it a while, I get the following SELinux alert.

# sealert -a /var/log/audit/audit.log
* Plugin catchall (17.1 confidence) suggests *
If you believe that ssl_crtd should be allowed read access on the
index.txt file by default. Then you should report this as a bug. You can
generate a local policy module to allow this access. Do allow this
access for now by executing:
# ausearch -c 'ssl_crtd' --raw | audit2allow -M my-sslcrtd
# semodule -i my-sslcrtd.pp

Unfortunately the suggested solution doesn't work, because the problem
reappears, and I get all sorts of errors in /var/log/squid/cache.log,
all due to ssl_crtd not being able to access stuff under /var/lib/ssl_db.

Any idea how I can solve this problem permanently ?

Cheers,

Niki

-- 
Microlinux - Solutions informatiques durables
7, place de l'église - 30730 Montpezat
Site : https://www.microlinux.fr
Blog : https://blog.microlinux.fr
Mail : i...@microlinux.fr
Tél. : 04 66 63 10 32
___
CentOS mailing list
CentOS@centos.org
https://lists.centos.org/mailman/listinfo/centos