Re: relayd.conf it's so confusing

2018-11-02 Thread Markus Rosjat

Hi again,


Am 02.11.2018 um 11:26 schrieb Markus Rosjat:
..  but also the match defined in the new defined protocol is still 
working. Thats something that shouldn't happen at all.



this seems to be resolved and was more or less browser related

--
Markus Rosjatfon: +49 351 8107224mail: ros...@ghweb.de

G+H Webservice GbR Gorzolla, Herrmann
Königsbrücker Str. 70, 01099 Dresden

http://www.ghweb.de
fon: +49 351 8107220   fax: +49 351 8107227

Bitte prüfen Sie, ob diese Mail wirklich ausgedruckt werden muss! Before you 
print it, think about your responsibility and commitment to the ENVIRONMENT



relayd.conf it's so confusing

2018-11-02 Thread Markus Rosjat

Hi all,

I have a relayd running that inspects the Host header of incoming 
traffic and then makes a decision to which server it should relay the 
traffic. so far so good but a few things don't add up after a few changes.


for example I have a protocol definition like so:

http protocol "httpproxy" {
match request quick header "Host" value "*domain1.tld" 
forward to 
match request quick header "Host" value "*domain2.tld" 
forward to 
}

and a relays like:

relay "www01proxy" {
listen on $gateway  port http
protocol "httpproxy"

forward to  port http
}

relay "www02proxy" {
   listen on $gateway  port http
   protocol "httpproxy"

   forward to   port http
  }

So this setup works but now it gets confusing if I add another protocol and 
relay to the above

http protocol "differenthttpproxy" {
match request quick header "Host" value "*domain3.tld" 
forward to 
}

relay "www03proxy" {
 listen on $gateway  port http
 protocol "differenthttpproxy"

 forward to   port http
}

now my relays 1 and 2 stop working, no traffic reaches the hosts, the order of 
the relays is

www03
www01
www02
 
in the config but it shouldn't be problem because the protocols used are diffrent. So coming to strange part two. I disabled the new relay and well the sites for relay 1 and 2 started to be reachable again but also the match defined in the new defined protocol is still working. Thats something that shouldn't happen at all.


what I did between the changes was checking sysntax and a

rcctl reload relayd

I am relucdent to do a restart because it happens to crash the VM. The VM is 
running 6.1 with all syspatches applied.

regards

--
Markus Rosjatfon: +49 351 8107224mail: ros...@ghweb.de

G+H Webservice GbR Gorzolla, Herrmann
Königsbrücker Str. 70, 01099 Dresden

http://www.ghweb.de
fon: +49 351 8107220   fax: +49 351 8107227

Bitte prüfen Sie, ob diese Mail wirklich ausgedruckt werden muss! Before you 
print it, think about your responsibility and commitment to the ENVIRONMENT