That’s pretty poor. Do you get an error message when trying to access it? Are you able to try a free proxy to use a different IP? Try this https://whatismyip.network/ instead of Myip.com.

@Actionhenk Nice. One other thing: I haven't looked at it but I assume if you need double quotes in a string like that in the future you can just escape the ones inside the string with \"

I believe that they are stored in definitely. SSH in and go to /var/lightsquid/report and you will see the report structure. It as a folder for every day, and inside each folder are the particulars for that day. The main report has options for showing year to date, so it must keep everything forever.

for other people with the same error. The problem was how mikrotik redirects the traffic. He used a dst nat rule which breaks the chain of trust. "Mangle" should be used

You can set the outgoing source address for requests which will effectively set the gateway used. That is done using the tcp_outgoing_address directive in the custom options before auth field. You would need static IP(s) to use that though. Steve

@douglas-filipe So after a big research I not found any solution... I tried to use HAProxy, without success too... I believe that reverse proxy packages doesn't work together with pf2ad solution or any other proxy auth solution. I'm right? I did new pfsense with no proxy auth and reverse proxy publish works great ! Best regards team!

@WD_Doug If none of the stock methods for reporting the user work for that, then you're going to have to hack something up. These might be useful: http://lightsquid.sourceforge.net/How%20It%20Work.html https://forum.netgate.com/topic/88886/lightsquid-not-showing-users/7 @mauricio2669 Not really, and even the report itself is built on the fly from several pieces so it's not easy to parse yourself.

@wesleylc1 Then it seems you have answered your question why the host header must be send in checks . HAProxy's health-checks by default do not send this header and your nginx configuration needs it to find the correct virtual host with the server_name configuration. The 'default virtual host' likely returns the 400 response status. Its nice to know that doing some research and actually understanding how things work you can succeed with fishing without needing other people to give you the fish, that will definitely help with future diagnosing of issues.

@kiokoman thanks!

Make a bug report at https://redmine.pfsense.org/projects/pfsense-packages

Yeah, that looks more like real data.

@gwaitsi Unfortunately I doesn't know, something wrong maybe with you CA. I done this before without any issue. It must process with success by cert install. No download needed. Can you try create new CA over Cert Manager and try again?

Some applications do not work well with PAC file. Are you using authentication ? Based on your config, it seems that you are bumping everything, tried splice all? Based on the problems I had, I found this: Some apps can work with PAC file, others not. Some apps can work with proxy authentication, like Kerberos for an exemple, others not. So, sometimes you will see Access Denied in Squid, because the apps like Pokerstars for an example are not carrying credentials to the proxy. Here, I have a Squid proxy with SSO, using Kerberos. Some apps don't work if I set Direct at the PAC file, however, they do work when I set a bypass like this: before_auth: acl whitelist dstdomain .pokerstars.com .dropbox.com <---- This allow everything to pokerstars.com and dropbox.com to go through the proxy without authentication auth_param negotiate program /usr/local/libexec/squid/negotiate_kerberos_auth -k /usr/local/etc/squid/mykeytab.keytab auth_param negotiate children 100 auth_param negotiate keep_alive on http_access allow whitelist <----- This allow whitelist before auth is required acl auth proxy_auth REQUIRED http_access deny !auth http_access allow auth

ACME/Let's Encrypt is the best thing to do here, assuming you have a public domain available you can leverage and a supported DNS provider.