Frequently Asked Questions
Installation and Configuration
Is it possible to use SysCP on a local server without a fixed IP? Im using a dial-up connection (e.g. DSL) and the domain is hosted at DynDNS.org.
There are two different ways of getting SysCP to work on this kind of system. If the server is connected behind a router using portforwarding, use the local FQDN as hostname and the local IP address. Is the server is directly connected to the Internet and only has a dynamic IP given by the provider, use * as IP address entry and the DynDNS.org hostname as SysCP hostname.
Where can I change the MySQL configuration for SysCP?
The file syscp/lib/userdata.inc.php contains all the settings for SysCP, as well as the MySQL configuration.
When and how often are the Webalizer statistics being created and where is the config file?
The Webalizer Statisics are being created every day at Midnight. There are no config-files, because webalizer is being started over the console as a programm and controlled by parameters.
Is it possible to use SysCP on a virtual server (vServer)?
SysCP can also be installed on a virtual server. Because of the shared kernel, there might be problems installing bind9. In that case, please contact your provider.
Why am I getting the error “Fatal error: Call to undefined function: mysql_error() in ...” or “You should install the PHP MySQL extension”?
Just restart Apache first. If this doesn't fix the Problem then do the following:
In php.ini the module mysql.so was not properly included. The following must be included in */etc/php4/apache/php.ini* and */etc/php4/syscpcron/php.ini*
extension=mysql.so
A restart of Apache is mandatory.
Is it possible to use Apache 2.0.x instead of 1.3.x?
Yes. Although SysCP currently depends on Apache 1.3, it has been proven that usage of Apache 2.0.x is possible without problems. However, we’d like to point out that you must use the tar.gz in that case.
Remark: In case of migrating to Apache2 just deactivate Apache1.3, due to dependencies it can’t be completely deinstalled.
I have installed everything according to the instructions, but still experience strange and faulty behaviour of various services.
Often Windows-Editors (e.g. Notepad.exe) are used for editing/saving configuration files and uploaded onto a Linux server. this practice leads into problems! Dos-Editors use <CR><LF> instead of <LF> (wich is used on Linux systems) at the end of a line. This applies in particular to all configuration files mentioned in the section “Server»Configuration” of your SysCP-Panel. Use Editors that can save in Unix/Linux-style.
I have installed Proftp according to the manuals. Why I still get the error “Failed binding to 0.0.0.0, port 21: Address already in use”?
You’ll find Inetd installed on most systems. In most cases inetd is running on these systems. Inetd takes care of the incoming packages on the ftp port and delegates them to the ftp-server. To do this delegation inetd has to bind port 21 on the ip-address 0.0.0.0. There is no way for another daemon to bind this port again.
To solve this problem you have to edit the inetd configuration file in /etc. Just comment out the line which tells inetd to listen on the ftp port and restart inetd. Now you have to restart proftpd and check the logfiles. If inetd was the reason why proftpd was not able to bind the port then there will be no error anymore. Otherwise there is another daemon which binds the ftp port.
This is an example how the line in your inetd configuration file could look like.
#ftp stream tcp nowait root /usr/sbin/tcpd /usr/sbin/proftpd
What entries in the sources.list are recommended for a trouble-free server-operation and SysCP?
The following entries in /etc/apt/sources.list should be sufficient for a normal operation of the main-services (e.g. Apache) to be controlled by SysCP:
deb ftp://ftp.debian.de/debian etch main deb-src ftp://ftp.debian.de/debian etch main deb ftp://ftp.debian.de/debian-non-US etch/non-US main deb-src ftp://ftp.debian.de/debian-non-US etch/non-US main deb http://security.debian.org/ etch/updates main deb-src http://security.debian.org/ etch/updates main deb http://debian.syscp.de/ etch/
Why I get the error “The following packages have unmet dependencies: syscp: Depends: Postfix but it is not going to be installed”?
You must first de-install exim (or other mail packages that were installed on your system and clashes with dependencies) Best this is done by using
apt-get install postfix
wich will force exim to de-install. As the next step use
apt-get install syscp
then dependencies should be ok.
As a customer, why can't I access a directory that was just created and instead get the error “The directory /xxx/ must exist, Please create it with your FTP client”?
The physical location where SysCP is installed, must have PHP safe mode off. In order to function properly, SysCP must also have access to the directories */var/www/syscp* and */var/kunden/webs*, assumed you are using the directory structure suggested by SysCP.
If you're running the SysCP-panel within a SysCP-administrated domain, just switch off Safemode and OpenBasedir for this domain.
Why does SysCP not create the user directories and/or the Apache vhosts?
The creation of user directories and Apache vhosts is done by the cronscript of SysCP. You should check first if the cronscript does run smoothly by typing:
/usr/bin/php5 -q /var/www/syscp/scripts/cron_tasks.php
If this gives an error, you should check your system again, and further should check why the cron daemon doesn’t send you any messages. (Note: PHP4 is not longer supported, you need to use php5 in order to use SysCP)
NOTE: In case you can't access SysCP after running the cronjob either create a single vhost-file for syscp or set vhostcontainer to 1 in the database (table panel_ipsandports, first entry).
Is it possible to use SysCP with FastCgi and Suexec or SuPHP?
Yes, several users reported success using SysCP with this configuration. A users contribution can be found here .
When trying to install/use SysCP mysql throws the error "Host '127.0.0.1' is not allowed to connect to this MySQL server"
Please check if localhost resolves to 127.0.0.1 and vice versa. If this fails, check your /etc/hosts and/or add the following line:
127.0.0.1 localhost
Don't forget to restart the mysql server afterwards.
How can I create an email box, that will receive all mails send to @domain.tld?
This kind of adress is called “catchall”. To add such an account, just switch on the flag *catchall* at the details. Per domain you can create one catchall-Address.
Why I can not receive eMails from outside the server?
It is already rejected during SMTP-negotiation. Please verify, if postfix has access to the MySQL-database(check MySQL-logs). Also verify, that only the latest configuration-files found in the Adminpanel are being used. Further keep in mind that postfix (to prevent hacking attacks) runs in a virtual “jail” Therefore it is neccessary to link the mysql-Socket into this jail:
ln /var/run/mysqld/mysqld.sock /var/spool/postfix/var/run/mysqld/mysqld.sock
But remember to renew the link after you restarted mysql. The easiest way to do this is to add a small check to the postfix startscript which takes care about the link.
What should I use as email username?
Just use the full email-address (e.g. user@domain.tld) as username for your mailclient.
I get the following error: “imaplogin: chdir webXY/<e-mailaddress>/: No such file or directory”. What is wrong?
This is a problem with your postfix installation and not of your cronscript as you may think. Your postfix has to create the maildirs and deliver the emails into them. If there is something wrong with this installation (perhaps an error on the database connection), postfix will not create these directories. Try to create a new email account and check the logs for the error on delivering the mail. There should be some useful information from postfix why the mail is not delivered to the maildir.
Nameserver/Domains
How can I use SysCP without a bind9?
Simply don’t install or start the bind9 daemon and use /bin/true as bind9 restart command in the admin settings of SysCP. Note: Due to Debian-dependency bind cannot completely de-installed.
I have created a subdomain. It also shows in the vhosts.conf. But I still can’t retrieve it.
First check if the domain is being resolved correctly and points to the IP-adress of the server. If not, the nameserver-configuration has to be changed. If you use the nameserver in combination with SysCP on the same server, first choice is to use the default.zone. In case you use a nameserver of your ISP, contact your provider for further assistance. It’s a good idea to have a wildcard (*) entry pointed onto your IP-adress.
Other Linux Distributions
Can I install and use SysCP on other Linux distributions?
Yes, several users reported success using SysCP on other distributions, e.g. Gentoo, SuSE, we even have reports about successfull installations on FreeBSD and NetBSD. Basically, a SysCP installation is possible on ANY server system with PHP and MySQL. The services SysCP depends on (POP3 server, mail transfer agent, FTP server) need to be compiled with MySQL support and configured to get their data from the SysCP tables. SysCP officially supportes Debian, however any porting to other systems or their HowTo’s in the forum are welcome. Please check out also our section Installation Manuals for updated installation manuals on other Distributions.
