In this post, I will describe a configuration to replicate a Linux website to another remote site. We will replicate everything required to restart the site on a remote server. Exact replication can be used on UNIX or Windows.
We can replicate between different hardware and different operating systems.
Although we are discussing a web server replication, the same article applies to any other server (physical or virtual) replication.
Introduction
In this example, the web server www.example.com will be replicated to www2.example.com.
Replication configurations are located under $edpcloud/etc.
We will replicate everything needed to run the server on the remote site, including the database.
What will we replicate
We will replicate only the following directories.
/var /home /etc /usr
We will create $edpcloud/etc/includes to instruct the replication software only to replicate what is in the includes file.
The includes file uses regular expressions.
bash-5.1# cat includes ^/home/.* ^/var/.* ^/etc/.* ^/usr/.*
What we will not replicate
We will use $edpcloud/etc/excludes to exclude specific file or directory patterns using regular expressions.
www# cat excludes ^/var/www/edxthemesite2/htdocs/myusage/.* ^/usr/local/licenses/.*
File replication configuration
We will create $edpcloud/etc/eddist.cfg, which controls replication. Here, the replication set l1 has a sender and a receiver section.
Sender www.example.com will replicate data to www2.example.com.
The receiver at www2.example.com will store replicated data in “/backup6t/sync3/%month%/%weekday%”.
<?xml version="1.0" encoding="UTF-8"?> <config name="ExapleCo" password="ThisISMyKnownPwdd122467$?" workers="2"> <link name="l1" > <sender hostname="www.example.com" /> <receiver hostname="www2.example.com" storepath="/backup6t/sync3/%month%/%weekday%" archivedir="/backup6t/syncarchive3/%date%" archive="1" archiveexcfile="archiveexcludes" /> </link> </config>
The receiver at www2.example.com will store replicated data in “/backup6t/sync3/%month%/%weekday%”.
Here, the “%variable%” will be resolved at run time as follows:
- %month%: Month of the year (01 to 12)
- %weekday%: Week Day (Monday to Sunday)
Notice that we also configured the archive directory to be archivedir=”/backup6t/syncarchive3/%date%” where:
%date% is in the format of YYYYMMDD. This means that any file that has changed will be archived in the archive directory, so we will have previous versions if we find the file under /backup6t.
Running replication
Replication can be run from a scheduler, command line, or in real time.
Running replication from the command line
To replicate /home:
edq -l l1 -r www2.example.com -n /home
Running replication in a schedule:
You can either use cron or edscheduler.cfg to replicate data:
Example:
www# cat edscheduler.cfg 0 7 * * * /usr/local/enduradata/edpcloud/bin/edq - l l1 -n /home 0 7 * * * /usr/local/enduradata/edpcloud/bin/edq - l l1 -n /var/www
Running replication in real-time
To replicate a directory in real-time, we create $edpcloud/etc/edfsmonitor.cfg
www# cat edfsmonitor.cfg /home /var/www
To restart replication services
#www edpcloud.sh startall
To stop data replication services
#www edpcloud.sh stopall
To dump and replicate the MySQL database used by our web server
DBDUMP=/tmp/wwwdbname.sql
/usr/local/bin/mysqldump --user=root --password='SOmePassWord" wwwdbname > $DBDUMP
/usr/bin/gzip $DBDUMP
edq -l l1 -n ${DBDUMP}.gz
Share this Post