Hi,

we are running a pound 2.6f with a patch v2 by Joe Gooch. We have a webservice
configured like this:

ListenHTTP
                Address XXX.XXX.XXX.86
                Port 80
                Service
                        Redirect "https://www.domain.de"
                End
End
ListenHTTPS
                Address xxx.xxx.xxx.86
                Port 443
                Cert "/etc/pki/tls/pound/domainname/domainname.pem"
                SSLAllowClientRenegotiation 0
                SSLHonorCipherOrder 1
                Ciphers "RC4-SHA:HIGH:!ADH:!SSLv2"
                Service
                        Backend
                                Address xxx.xxx.xxx.xxx
                                Port 80
                        End

                        Session
                                Type COOKIE
                                ID "userblbbla"
                                TTL 360
                        End
                End
End
#############################
When we do a http://www.domain.de/page1.html on
our Site, we get a redirect to https://www.domain.de, which is as expected,
but we are missing the "page1.html" part. (yes the page.html part can change
very often)


We can archive this in a normal termination on a apache with something like
this:
 RewriteCond %{SERVER_PORT} !443
 RewriteRule ^/(.*)$ https://www.domain.com/$1 [R=301,L]


Is it possible to let the pound act like this ?


Kind regards


fatcharly


[...]

You'd want the RedirectAppend option provided here:
https://github.com/goochjj/pound/commit/6987fab10d7ee47d0b46501520a934839c3085db

It may be easier to use the branch I staged for prospective 2.7... you can do
that with:

git clone git://github.com/goochjj/pound.git -b stage_for_upstream/v2.7
cd pound
configure/compile as usual

The pound.8 man page in the repo will illustrate how to use Redirect vs
RedirectAppend vs RedirectDynamic.

Joe


[...]

Hi Joe,

when I try to compile this "git"-pund, I run into this error:

[root(at)pilotpound pound]# make
gcc -DF_CONF=\"/etc/pound.cfg\" -DVERSION=\""2.6"\" -DC_SSL=\"""\"
-DC_T_RSA=\""0"\" -DC_MAXBUF=\""0"\" -DC_OWNER=\"""\" -DC_GROUP=\"""\"
-DC_SUPER=\""0"\" -DC_CERT1L=\"""\" -g -O2 -pthread -DUPER -DNEED_STACK
-DHAVE_SYSLOG_H=1 -pthread  -D_REENTRANT -D_THREAD_SAFE -Wstrict-prototypes
-Wno-unused-result -pipe   -c -o config.o config.c
config.c:83: error: expected â=â, â,â, â;â, âasmâ or â__attribute__â
before âCNNameâ
config.c: In function âparse_HTTPSâ:
config.c:1159: error: âCNNameâ undeclared (first use in this function)
config.c:1159: error: (Each undeclared identifier is reported only once
config.c:1159: error: for each function it appears in.)
config.c: In function âparse_fileâ:
config.c:1396: error: âThreadModelâ undeclared (first use in this function)
config.c: In function âconfig_parseâ:
config.c:1524: error: âThreadModelâ undeclared (first use in this function)
config.c:1591: error: âCNNameâ undeclared (first use in this function)
At top level:
cc1: warning: unrecognized command line option "-Wno-unused-result"
make: *** [config.o] Error 1
[root(at)pilotpound pound]#


Any Suggestions are welcome

Kind regards

fatcharly


-------- Original-Nachricht --------[...]
[...]
[...]

Good catch!  Search for CNName in config.c, there should be a comma after
Threadmodel before cnname.

Let me know if that fixes it... If no-unused-result is still an issue you can
take that bit out of configure.
Joe
[...]

On Wed, 2012-04-04 at 13:42 +0200, fatcharly(at)gmx.de wrote:[...]

1. Please make sure you run with the final 2.6
2. If the behaviour persists, it is a bug (the redirect should be to
https://www.domain.de/page.html).
Let me know and we'll look into it.[...]

Interesting, apparently Robert implemented this in 2.1.6. (2.0 didn't have the
Append feature)  I never noticed.

My apologies, stock 2.6 should be working for you.

Try using LogLevel 2, it should issue a REDIRECT log message you can use to see
exactly what the client is being redirected to.

And ensure there's no trailing / in the URL you're providing on the Redirect
directive.

Let me know if you still have problems with stock 2.6! (or 2.7a)
Joe
[...]

-------- Original-Nachricht --------[...]
[...]

I´m sorry but it seems my browser was acting weired. It works like a charm on
another machine. So everything is fine for me.

Thanks for your and joe´s support

Kind regards

fatcharly

 
[...]
[...]