Using MyBackup

Return to Index

Database Backup

With MyBackup, you can schedule complete database backups, selective backups of specific tables within a database, or multiple database backups. In the screenshot below, I've specified a few tables in the "gpix" database that I'd like to have backed up.

The SQL options are essentially the same as those that you would find in phpMyAdmin or phpFlashMyAdmin. You may, for example, add "drop table if exists" commands before the exported SQL, to help facilitate database restorations. Please note that MyBackup does NOT restore your database, it just helps back it up. Future versions of MyBackup may have restore capabilities.

When you initiate a backup, the system will create a new "job" in the /cronjobs/ folder, and it will execute this job:

Thus, initiating the backup, performs the backup, and saves the configuration of your backup so that it can be quickly and easily repeated in the future in the form of a cronjob.

Web Files Backup

Backup is great if you'd like to schedule backups of specific files and/or folders on your server. For example, suppose that you have a folder that many users upload files to on a regular basis. Since you may not always have a copy of those files (they might be uploaded by users via a web form, for example), you should schedule regular backups of the folder.

By default, MyBackup uses your document root folder as the root to determine what folder to make available for backup. However, you can change that in config.php. Locate the "FILE_DISPLAY_ROOT" constant in config.php, and extend the file path shown. Please be sure to add the final "/" at the end of the path. In this case, instead of reading web files from my document root folder (/server/www/), I chose to read from the "gpix" subfolder.

define("FILE_DISPLAY_ROOT","c:/server/www/gpix/");

Cronjobs

When you create/execute a backup, your backup is automatically saved in the /cronjobs/ folder. This permits you to RE-RUN your backup any time that you want to. The PHP file in the /cronjobs/ folder is simply an HTML form with the options that you specified. When the PHP file is directly accessed, the backup "job" is executed - i.e., the backup of your database or web files is made and then emailed, or ftp'd, or saved to your server, or whatever you wanted it to do.

The PHP file is named with the month, date, year, hour, minute, and second that the backup job was created. In this case, if I were to directly access this URL, the backup job would re-start:

http://www.tufat.com/backup/cronjobs/July_14_2006_5_53_43_pm.php

Of course, this backup doesn't actually exists, because I don't want just anyone to receive backups of the TUFaT.com database! :)

This opens up an important tool for us: automatic backups using Unix and Linux "cronjobs". If you have a Unix or Linux server available to you, you should be able to setup a "cronjob", which can remotely execute a PHP script at a regular interval.

For example, using CPanel, you can create a cronjob that looks something like this:

The "Command" actually looks like:

GET http://......php > /dev/null

In order to use this command syntax, you must have the Lynx utility installed on your server. Of course, if you have direct shell access to a server, you can manually create cronjobs by manually editing your crontab file.

The PHP files in the /cronjobs/ folder are simply HTML forms with the specific backup settings saved in the form of an HTML form. When the PHP file is accessed, the form is auto-submitted and the backup is performed.