cherrun's questions

I have two domains:

foo.com and bar.com

Only foo.com is using Google Apps and thus can send emails. I now have a service running on bar.com\application that uses the email [email protected] to send emails to others. However, all the mails get delivered to the spam folder of the recipient. I checked the header and it says

spf=neutral (google.com: 2607:f8b0:4001:c02::2ef is neither permitted nor denied by best guess record for domain of [email protected]

Therefore I would like to set up a spf record. Now I don't know for which domain I have to set the spf. Do I have to set it for foo.com, because the mails get send with that domain or do I have to set it in bar.com, because that's where the mail get send from?

In addition, I heard setting up a DKIM entry in the nameserver is recommended as well, but does this make sense in my case? As far as I understand, DKIM verifies that a mail is in fact coming from the domain/IP, which uses the mail service (foo.com in my case). That, however, would mean sending from bar.com using [email protected] would be bad, because it didn't get send from bar.com?

Is it somehow possible to extend one's webspace by connecting to a server which has more space?

Lets say I have a remote server with a small disk, and a local server with plenty of free disk space. Can I somehow permanently mount the space on my local server on my remote server?

How would I set up Dovecot and Postfix to use a different user database than passwd for the email accounts? I'd would like to have separate accounts for user, who can actually log in via ssh, and user who have a mail address, although those might be the same person.

Lets say I have a account A, who can connect to the server via ssh. Now I want to have an [email protected] mail address, but with a different password than the A account for ssh. The emails should nonetheless being saved in A's home folder.

How can this be accomplished?

Until now I have a directory structure defined in /etc/skel so that each ssh Account will have the mail directory automatically included. Then I set up the mail addresses via Postfix within the virtual file and then running 'postmap'.

I want to run multiple domains on my server (only two at the moment). I already changed both nameservers to point to my server.

In the apache2.conf file I enabled "included sites-enabled".

My ports.conf file looks as follows

NameVirtualHost *:80
Listen 80
Listen 443

<VirtualHost xx.xx.xx.xx:443>
    SSLEngine On
    SSLCertificateFile /etc/apache2/ssl/apache.pem
    SSLCertificateKeyFile /etc/apache2/ssl/apache.key

    ServerAdmin [email protected]
    ServerName firstDomain.com
    DocumentRoot /var/www/
    ErrorLog /var/log/error.log
    CustomLog /var/log/access.log combined
</VirtualHost>

This is the setting for my main page.

My httpd.conf file looks like this

ServerName firstDomain.com

<VirtualHost xx.xx.xx.xx:80>
ServerAdmin [email protected]
ServerName firstDomain.com
DocumentRoot /var/www/
ErrorLog /var/log/error.log
    CustomLog /var/log/access.log combined
</VirtualHost>

<VirtualHost *:80>
    ServerName secondDomain.com
    ServerAlias *.secondDomain.com
    DocumentRoot /home/user/www/secondDomain.com/htdocs
</VirtualHost>

Options -Indexes All FollowSymLinks MultiViews

Last but not least my "secondDomain" file in "sites-avaiable"

<VirtualHost secondDomain.com>
    ServerAdmin [email protected]
    ServerName  secondDomain.com

    # Indexes + Directory Root.
    DirectoryIndex index.html
    DocumentRoot /home/user/www/secondDomain/htdocs/

    # CGI Directory
    ScriptAlias /cgi-bin/ /home/user/www/seoncDomain/cgi-bin/
    <Location /cgi-bin>
            Options +ExecCGI
    </Location>


    # Logfiles
    ErrorLog  /home/user/www/secondDomain/logs/error.log
    CustomLog /home/user/www/secondDomain/logs/access.log combined
</VirtualHost>

I also get an error message, everytime I restart the apache server, saying that VirtualHosts xx.xx.xx.xx:80 overlaps with VirtualHosts xx.xx.xx.xx:80.

When I call my secondDomain.com only the firstDomain.com pops up. It seems like the mapping is wrong.

apachectl -t output:

[warn] VirtualHost xx.xx.xx.xx:80 overlaps with VirtualHost xx.xx.xx.xx:80, the first has   precedence, perhaps you need a NameVirtualHost directive
Syntax OK

apache2ctl -S output:

[Wed Feb 29 02:05:17 2012] [warn] VirtualHost xx.xx.xx.xx:80 overlaps with VirtualHost xx.xx.xx.xx:80, the first has precedence, perhaps you need a NameVirtualHost directive
VirtualHost configuration:
127.0.0.1:*           secondDomain.com (/etc/apache2/sites-enabled/secondDomain.com:1)
xx.xx.xx.xx:443       firstDomain.com (/etc/apache2/ports.conf:12)
xx.xx.xx.xx:80        firstDomain.com (/etc/apache2/httpd.conf:3)
wildcard NameVirtualHosts and _default_ servers:
*:80                   is a NameVirtualHost
     default server secondDomain.com (/etc/apache2/httpd.conf:11)
     port 80 namevhost secondDomain.com (/etc/apache2/httpd.conf:11)
     port 80 namevhost firstDomain.com (/etc/apache2/sites-enabled/000-default:1)
     port 80 namevhost firstDomain.com (/etc/apache2/apache2.conf:241)
     port 80 namevhost secondDomain.com (/etc/apache2/apache2.conf:249)
Syntax OK

Wordpress is installed through apt-get. I have a symlink in /var/www/ which points to /usr/share/wordpress and activated in /etc/apache2/sites-available and a2ensite.

I want to have a secure connection, when I log into my webmail, phpMyAdmin, etc.

Therefore I signed my own SSL certificates with OpenSSL and told Apache to listen on port 443.

Is this in fact secure? Are all my passwords really sent through a safe and secure layer? What difference does it make, if I buy an SSL certificate from Verisign or sign my own one? At the end of the day, all data will be on my server anyway. So what is the big difference?

Possible Duplicate:
How to know what user belongs to what ip on debian

Is there a way to see a list of all current logged in user on a Debian server? Sometimes it is very useful to do so, e.g. before updating certain plugins etc.

Any software or built-in features? All users will log in through ssh.

Where lies the config file to change the output of the access.log file? I want to remove the website the user was browsing and format the time so it's human readable.

I have a virtual server in the US. When I first set it up I was able to use dynamic port forwarding with

ssh -D PORTNUMBER serveraddress

After that I could set up Firefox to use that as SOCKS5 proxy. I don't know what changed, but for some reason that doesn't work anymore.

Is there a place, where it is set if tunneling is allowed or not? I tried ssh_config file, there was in fact a

Tunnel yes/no

option, but even after switching it to yes and restarting, it didn't change anything.

Where should I look?

EDIT (01:41 GMT+1): The output with -vvv options is rather long, therefore uploaded it to pastebin. I hopefully removed all sensitive data: http://pastebin.com/r4BVC5F1

EDIT (02:13 GMT+1): New output file: http://pastebin.com/VvPVv0ff

I added the "PermitTunnel" option in my config file, now I can log in with -vvv active. However the proxy still won't work. I havent changed on settings on the client side, at least none that I know of. And yes, I 'm using Mac OSX.

EDIT (09:45 GMT+1): Could this have anything to do with messed up hostnames? Or maybe dante or squid configurations intereferring? I seriously don't know, why it's not working anymore...

LAST EDIT: OK, I'm just plainly stupid! Seriously. Everything worked fine. The problem was that I typed in the IP address of my server into the SOCKS server settings. Which can't possibly work. localhost ist of course the right answer. Darn... wasted so much time... Thanks everyone!

I have the following .htaccess file:

RewriteEngine   on
RewriteBase     /

RewriteRule     ^(/?)dir/?$            $1dir/subdir           [L]
RewriteRule     ^(/?)subdir2/?$        $1dir/subdir/subdir2   [L]

The redirecting works fine, if I type in the IP of the domain, but typing in the domainname/dir doesn't redirect me at all, instead shows me the content of the directory.

Another question: As I said, the redirecting with the IP works fine. However, although the 'R' flag is not set, the url changes in the url box. IP/dir changes into IP/dir/subdir. Why?

EDIT (11:20 GMT+1): Not quite sure if it's of any relevance, but before today the server didn't have a domain. I just bought a domain and set up all the nameserver pointing to the server up. Since I can access the page and it shows me the content of the /dir folder, I assume that I didn't mess up anything there.

EDIT (12:51 GMT+1): Output from apachectl -S:

apache2: Could not reliably determine the server's fully qualified domain name, using 0.0.10.188 for ServerName
VirtualHost configuration:
wildcard NameVirtualHosts and _default_ servers:
*:80                   is a NameVirtualHost
     default server 0.0.10.188 (/etc/apache2/sites-enabled/000-default:1)
     port 80 namevhost 0.0.10.188 (/etc/apache2/sites-enabled/000-default:1)
Syntax OK

EDIT (19:26 GMT+1): Found the problem. The problem was that my server name was not set and I had duplicates in ports.conf and httpd.conf, both where "listening". Thanks for pointing out the "apachectl -S" command. One mysterious problem left: Why is the url changing, although no R flag is set. Forever a mystery...

I just set up my dante socks proxy and it works just fine without authentication. Now I want only certain users to be able to use it tho. Preferably only those users that are in the passwd. What would you guys recommend me using? Can I somehow use PAM and link it to my passwd? What other authentication is there and how would they work? For example would an authentication window pop up, before the user would be able to browse?