Open phpMyAdmin or mysql command line client, choose database pma and execute query:
USE psa; SELECT REPLACE( sys_users.home, '/var/www/vhosts/', '' ) AS domain, sys_users.login, accounts.password FROM sys_users LEFT JOIN accounts ON sys_users.account_id = accounts.id ORDER BY sys_users.home ASC
It will show you all FTP users from Plesk
This article describes upgrade CentOS 5.5 from i386 arch to x86_64 arch.
First of all, upgrade kernel and boot loader from CD. For my system this operation was done by hosting provider.
After kernel upgrade system have a lot of i386/i686 packages. It isn’t possible to upgrade each package manually. For example, my CentOS had 300+ packages.
To continue you need to install yum
The algorithm of upgrading is simple:
But it can be complicated, because any packages uses the same paths or something else and you need manually to install RPMs.
Let’s begin.
Install yum plugin yum-downloadonly:
yum install yum-allowdowngrade
Then take a list of packages for upgrade:
rpm -qa --qf "%{name}.%{arch}\n" | grep -E "i386|i686"
Copy this list into your text editor and replace “.i386” and “.i686” with “.x86_64”. Also replace all returns with spaces. You will take something like this:
glibc.x86_64 readline.x86_64 ncurses.x86_64
Now download this packages RPMs:
yum install --downloadonly --downloaddir=/root/rpms/ <packages>
where <packages> are in my example glibc.x86_64 readline.x86_64 ncurses.x86_64
Of course directory/root/rpms/ must exist.
When yum finishes to download hundreds of packets, chdir to /root/rpms/ and execute
rpm -i --force *.rpm
rpm will force installation of all downloaded packages.
After installation you need come back to your i386/i686 packages list and uninstall them.
yum erase <packages>
where <packages> are in my example glibc.i686 readline.i386 ncurses.i386
That’s all!
Plesk license key is located:
1. Parallels Panel 7.x – 8.x:
/etc/psa/psa.key
2. Parallels Panel 9+:
/etc/sw/keys/keys/key*
There is 2 steps.
First of all, add to /etc/proftpd.include followed code:
TimeoutIdle 900 TimeoutNoTransfer 900 PassivePorts 49152 65534 TimesGMT Off
Then test the iptables nat helper:
modprobe ip_conntrack_ftp
If there is no errors, add to directive IPTABLES_MODULES in /etc/sysconfig/iptables-config module ip_conntrack_ftp:
IPTABLES_MODULES="ip_conntrack_ftp"
If you need to load two or more helpers, separate them with spaces because the iptables script does a for-in on the IPTABLES_MODULES directive.
When you have a newly installed Plesk, it’s not enought to run PHP-scripts in CGI. Unfortunately.
You need manually copy suexec file from Plesk binaries file folder to system binaries folder.
In common cases it does command:
# cp /usr/local/psa/suexec/psa-suexec /usr/sbin/suexec
If cp cannot find any of files, search they manually
# locate suexec
Important! It’s recommended to make a copy of system’s suexec file
# cp /usr/sbin/suexec /usr/sbin/suexec_old
I’ve needed to allow for my domain only hosting without mail service. Mail server will be remote. I’ve changed MX-records to remote server IP (or hostname), but it’s not enough. Also this domain must be removed from mail daemon (Qmail, Postfix or other) config files.
There is no possibility to disable mail service for domain via Small business panel. You can only do it via command line.
#/usr/local/psa/bin/domain -u <domain> -mail_service false
Additionally, if you want to set option “What to do with mail sent to nonexistent users” to “Forward to address”, you need than manually edit
/var/spool/postfix/plesk/virtual.db
with command
postmap -d @<domain> hash:virtual
By default in Plesk you can view AWstats statistics for each domain separately at http://<yourdomain>/webstat . And you must type here FTP password for the domain. It’s not comfortably when your server holds much more than one domain.
So, our task is to create superadmin-access for AWStats for viewing all domains statistics with only one password.
First of all, find you awstats.pl file. It must be located in cgi-bin directory which lies near your vhosts. In my case it’s /srv/www/cgi-bin
Put here 2 files:
.htaccess:
AuthUserFile /srv/www/cgi-bin/.htpasswd AuthName "AWStats" AuthType Basic require valid-user
index.pl:
#!/usr/bin/perl
use strict;
use warnings;
use CGI qw(:standard);
print header;
print start_html(-title => 'AWStats');
opendir (DIRH, '/usr/local/psa/etc/awstats') || die; # path depends on your system
print '<div align="center">';
my $file; my @files;
# awstats config files are named like awstats.<yourdomain>-http.conf
# if you need to see also ftp etc, modify my code
while ($file = readdir DIRH) {
$_ = $file; # my perl doesn't works without assignment result of readdir
next unless m/\-http\.conf$/;
push @files, $file;
}
for $file (sort @files) {
$_ = $file;
s/^awstats\.//;
s/\.conf//;
print "<a target='_blank' href='awstats.pl?config=",$_,"'>";
s/\-http//;
print;
print '</a><br>';
}
print "</div>";
Don’t forget to chmod +x index.pl
To finish with this dir, create .hpasswd file:
htpasswd -c .htpasswd <username>
At last you need to config Apache web-server.
Put in /etc/apache2/conf.d/awstats.conf (or create a new one):
ScriptAlias /awstats-full /srv/www/cgi-bin Alias /awstats-icon /usr/share/apache2/icons/awstats <Directory /srv/www/cgi-bin> AllowOverride All </Directory>
Please check all paths, they are depend on your system. For exampe, apache config dir may be /etc/httpd/
Finally, restart Apache graceful (recommended)
apachectl -k graceful
or restart fully
apachectl -k restart
Now you can see all domain statistics on http://<your ip>/awstats-full/index.pl
I’ve found in Net script, which only backups databases into separate files and modified it. I’ve added gzip compression and backup by weekday. So, you can cron this job daily and get backups for each weekday. It’s very flexible backup algorithm.
Script is below:
#!/bin/bash # This script backups every MySQL database to its own file #Some variables you can set how you like USER='<possible root>' PASSWORD='<password>' DAYOFWEEK=`/bin/date +"%w"` OUTPUTDIR="/usr/backup/mysql/$DAYOFWEEK" # backup dir MYSQLDUMP='/usr/local/bin/mysqldump' # path to mysqldump, may be /usr/bin/mysqldump MYSQL='/usr/local/bin/mysql' # path to mysql, may be /usr/bin/mysql #Clean up any old backups rm -f $OUTPUTDIR/* #Get a list of databases names except the system one databases=`$MYSQL --user=$USER --password=$PASSWORD -e 'SHOW DATABASES;' | grep -Ev '(Database|information_schema)'` #Dump each database in turn and compress the output with gzip for db in $databases; do $MYSQLDUMP --opt --hex-blob --force --user=$USER --password=$PASSWORD $db | gzip > $OUTPUTDIR/$db.gz done
Download script mysql_backup.sh (800 kB)
Function below allow you to get count of Google backlinks for requested page
function getGoogleOuterLinksCount($url) {
$domain = preg_replace('/^www\./','',array_shift(split('/',$url)));
/* I strongly recommend you to use www.google.<you country zone> version of Google */
$html = "http://www.google.com/search?q=link:".$domain."&hl=en";
$content = getRemoteFile($html);
if (preg_match('/<div id=resultStats>(\d+) results/i',$content,$arr)) {
$t = $arr[1];
$t = str_replace(' ','',$t);
$t = str_replace(',','',$t);
return (int)$t;
} else
return 0;
}
function getRemoteFile($url) {
$ch = curl_init($url);
curl_setopt($ch, CURLOPT_HEADER, 1);
curl_setopt($ch, CURLOPT_RETURNTRANSFER, 1);
$data = curl_exec($ch);
if(curl_errno($ch)) {
echo "Error! Error code:".curl_errno($ch)." Error:".curl_error($ch);
curl_close($ch);
return false;
}
list($header, $data) = preg_split("/\r?\n\r?\n/", $data, 2);
return $data;
}
The function below allow you to get the count of pages indexed by Google.
function getGoogleIndexCount($url) {
$html = "http://www.google.com/search?hl=en&safe=off&q=site%3A".$url."&btnG=Search"; // I strongly recommend you to use www.google.<you country zone> version of Google
$content = getRemoteFile($html);
if (preg_match('/<div id=resultStats>(\d+) results/i',$content,$arr)) {
$t = $arr[1];
$t = str_replace(' ','',$t);
$t = str_replace(',','',$t);
return (int)$t;
} else return 0;
}
function getRemoteFile($url) {
$ch = curl_init($url);
curl_setopt($ch, CURLOPT_HEADER, 1);
curl_setopt($ch, CURLOPT_RETURNTRANSFER, 1);
$data = curl_exec($ch);
if(curl_errno($ch)) {
echo "Error! Error code:".curl_errno($ch)." Error:".curl_error($ch);
curl_close($ch);
return false;
}
list($header, $data) = preg_split("/\r?\n\r?\n/", $data, 2);
return $data;
}