After I purchased the NAS box to place at home for my work data (there is a separate one for family data, they do however backup to each other but I will cover that in another post) I decommissioned my old Windows Server 2008 R2 box.
This box, however, did do a multitude of things that were controlled via scheduled tasks and scripts that I have now moved to the Synology. Chief amongst this was the backup for several websites for “just for when” something goes wrong.
There were several bits of software in the implementation of this task, these were (are);
- wget (Windows Version) – Command line utility for downloading files, whilst there are other options, this was quick and simple, exactly what I needed
- FTPSync (CyberKiko) – a Great little piece of software, can display a GUI showing sync progress which is useful for troubleshooting or runs in a silent mode with no GUI. It utilises simple ini text files for configuration (it encrypts the password) making it easy to configure and it has many options for doing this configuration
- DeleteMe (CyberKiko) – Simple file removal tool, give it a folder (it can have multiple set up) and a maximum age of the files in that folder and it will remove anything older than that.
- 7Zip (Command Line Version) – Command Line zip archive creation utility, what more is there to say
- Custom PHP DB Export Scripts – Custom PHP scripts that pulls the database(s) out of MySQL and zips it up. This was originally run with a CRON job, but I found it easier to use wget to pull the trigger file when I wanted the backup was then created, then pull the file itself, then pull a delete trigger
That’s it for the software I use but what about the backup process itself? For each of the sites, I need to backup the custom PHP scripts were configured on the server. Then a custom batch file containing a bunch of commands (or should that be a batch of commands) to download and archive the files.
The batch file had the following segments in it to achieve the end goal;
- Check if backup is still running from previous attempt (Utilizes blank text file that is created at start of script and then removed at end)
- If it is running, skip the script and go to the end of the file
- If a backup job is not running, create the file locking out other jobs.
- Run cleanup of old files
- If an existing backup directory for today exists (due to a failed backup job most likely), remove it and create a new one
- Start logging output to a log file
- Start Repeating Process (Repeats once for each site that is being backed up)
- Generate Database Backup
- Retrieve Database Backup
- Remove Database Backup to the long term storage folder
- Rename Database Backup File
- Move Database Backup File to Storage Location
- Sync (utilizing FTPSync) the sites directories
- Remove Existing zipped backup file of the site’s files and directories if it exists
- Zip folder structure and files for the website put the ZIP file in the long term storage folder
- Copy Backup Complete information to log file
- Remove Process Lock File
To download the batch files, click here
Reasonably simple, to add a new site, copy and paste a previous one, update a few details and off you go.
Now I realize that some of this is perhaps not the best or most secure way to achieve a goal (specifically how I was handling the database) but it was quick, easy and it worked. I could have also made the whole process more efficient by using config a files and a for loop, but well I didn’t