I’m running Virtualmin on 2 Centos 7 servers and both are producing the same backup file format.
I’m opening the backup file in 7zip on Windows 7.
Regardless of which backup format I choose (single archive or one file old/new) I get a tar which opens to the expected structure which includes the domain_dir. I can open this but it consists entirely of numbered folders. It is not representative of the original folder structure although opening the numbered folders shows pieces of the original. For example, here are a couple of paths from a backup, neither of which contain the expected content:
Virtualmin includes some additional metadata in its backups so that it can restore the domain and all of its various configuration. Because that data spans many directories (/etc, /var, /home) it isn’t as simple as untarring the file. All of your data should be there, and reasonably easy to pull out, even without restoring it into Virtualmin. Our goal was to not make it too obscure or obfuscated…but, it’s slightly more complex than a simple tarball containing home.
Thanks very much for the answer Joe but if you can offer some further clarification I’d really appreciate it.
I’ve attached a backup file that I created using the settings below, you’ll need to remove the .txt extension to open it. I’ve also attached a screenshot of 7zip showing the content of the backup archive as I see it:
Only selected
Backup all features
Local file or directory (set to /tmp)
One file per server
Full (all files)
The structure of my public_html folder is as follows:
awstats-icon (sym link)
awstatsicon (sym link)
icon (folder created by system)
stats (folder created by system)
css (folder)
css > style.css
images (folder)
images > test.png
index.html
So you can see I have a very basic structure with 1 style sheet, 1 image and an index page. However, the backup contains just numbered folders. As you say, my data should be there but it’s impractical, even with such a simple structure to find it, and would be virtually impossible with a more complex site to restore it in the original format.
So my question again is, is this the expected format or am I seeing something I shouldn’t be? Your comments of ‘your data should be reasonably easy to pull out’ and ‘Our goal was to not make it too obscure or obfuscated’ make me think I’m not seeing what you think I’m seeing.
Hi, to open tar files etc… just install peazip onto your windows machine or 7zip… both open source or use anything else instead of windows os like ios or linux or unix