Does this also auto solve for IPv6 connections changing it to just 

http_port 3128
https_port 3129??

> On Jul 12, 2024, at 04:57, Amos Jeffries <squ...@treenet.co.nz> wrote:
> 
> On 12/07/24 11:50, Jonathan Lee wrote:
>>> I recommend changing your main port to this:
>>> 
>>>   http_port 3128 ssl-bump ....
>> This is set to this when it processes
>> http_port 192.168.1.1:3128 ssl-bump generate-host-certificates=on 
>> dynamic_cert_mem_cache_size=20MB cert=/usr/local/etc/squid/serverkey.pem 
>> cafile=/usr/local/share/certs/ca-root-nss.crt capath=/usr/local/share/certs/ 
>> cipher=EECDH+ECDSA+AESGCM:EECDH+aRSA+AESGCM:EECDH+ECDSA+SHA384:EECDH+ECDSA+SHA256:EECDH+aRSA+SHA384:EECDH+aRSA+SHA256:EECDH+aRSA+RC4:EECDH:EDH+aRSA:!RC4:!aNULL:!eNULL:!LOW:!3DES:!SHA1:!MD5:!EXP:!PSK:!SRP:!DSS
>>  tls-dh=prime256v1:/etc/dh-parameters.2048 
>> options=NO_SSLv3,NO_TLSv1,SINGLE_DH_USE,SINGLE_ECDH_USE
> 
> The key thing here was the removal of the IP address. So that Squid received 
> both the 192.168.*.* and the 127.0.0.* traffic without needing separate 
> http_port lines.
> 
> 
> 
>>> and receiving the intercepted traffic on:
>>> 
>>>  http_port 3129 intercept ssl-bump …
>> Do you mean https?
> 
> Sorry. I missed that you had an https_port using 3129 already.
> 
> 
> 
>> https_port 127.0.0.1:3129 intercept ssl-bump generate-host-certificates=on 
>> dynamic_cert_mem_cache_size=20MB cert=/usr/local/etc/squid/serverkey.pem 
>> cafile=/usr/local/share/certs/ca-root-nss.crt capath=/usr/local/share/certs/ 
>> cipher=EECDH+ECDSA+AESGCM:EECDH+aRSA+AESGCM:EECDH+ECDSA+SHA384:EECDH+ECDSA+SHA256:EECDH+aRSA+SHA384:EECDH+aRSA+SHA256:EECDH+aRSA+RC4:EECDH:EDH+aRSA:!RC4:!aNULL:!eNULL:!LOW:!3DES:!SHA1:!MD5:!EXP:!PSK:!SRP:!DSS
>>  tls-dh=prime256v1:/etc/dh-parameters.2048 
>> options=NO_SSLv3,NO_TLSv1,SINGLE_DH_USE,SINGLE_ECDH_USE
>> Https uses that port 3129
>> What should I adapt
>> http_port
>> https_port?
> 
> Both.
> 
> FYI, there are two issues:
> 
> 1) listening on IP 127.0.0.1. Inside the OS there are different devices for 
> localhost (lo) and WAN (eg. eth0). NAT is problematic already without 
> introducing any tricky behaviours from bridging those "private" (lo) and 
> "public" WAN devices.
> 
> The simplest solution is just not to put any IP address on the squid.conf 
> *port line(s) with intercept options. The OS will select one appropriate for 
> whatever device and tell Squid on a per-connection basis.
> 
> The more difficult way is to put one of the machines "global" (WAN or LAN) IP 
> addresses. In your case 192.168.1.1. With most connections being from the LAN 
> that minimizes the possible problems.
> 
> 
> 2) listening on a well-known proxy port 3128 for intercepted traffic.
> 
> There is malware in existence that scans for at least port 3128 (likely 1080, 
> 8080 etc common proxy ports) being used by proxies like yours and abuses 
> them. As a result at least one popular antivirus network scanner (from Trend) 
> does the same scan to detect insecure proxies.
> 
> The worst thing about this situation is that the NAT very effectively hides 
> the malware. So it is extremely hard to see whether it is happening to you.
> 
> 
> I am not sure what UI you are using to show those firewall rules in your 
> other email. However the one that had ALLOW for the port range 3128-3129 
> worries me. AFAIK that should only be for 3128 and a separate rule somewhere 
> else to drop the intercepted port 3129 traffic pre-NAT.
> 
> 
> HTH
> Amos
> _______________________________________________
> squid-users mailing list
> squid-users@lists.squid-cache.org <mailto:squid-users@lists.squid-cache.org>
> https://lists.squid-cache.org/listinfo/squid-users

_______________________________________________
squid-users mailing list
squid-users@lists.squid-cache.org
https://lists.squid-cache.org/listinfo/squid-users

Reply via email to