Hi people,

I have about 150 mb worth of sites on my RaQ3 and I'd rather like to zip and download them so they are actually unzippable on my pc.

Is that possible? I'm not too worried about the time it takes or the bandwidth involved, I'd just like an ultra up-to-date backup which i can use to amend and re-upload the odd file.

Any ideas?

Simple. su to root.

tar -cvfp /home/sites.tar /home/sites/*

You'll get a nice tarfile @ /home/sites.tar with everything, including the symlinks.

If you need any further help just mail me!

Best,
Drew

Thanks for the reply. This seems to make a 1GB+ file which is named "p", rather than sites.tar

It should not. Did you copy/paste the code I entered here?

Drew

Try: tar -pcvf /home/sites.tar /home/sites/*

Originally posted by horoscopes2000
I have about 150 mb worth of sites on my RaQ3 and I'd rather like to zip and download them so they are actually unzippable on my pc.

I'm "backing up" >150Megs worth of files via mirrordir, which runs nightly (via Cron) on a (Linux) PC. Other than typing "pine" on the Linux machine every morning to check the mirrordir "reports" from the nightly cron jobs), the backup process is automated.

Mirrordir duplicates file-systems in every detail, even correctly recreating hardlinks, devices and access times. Mirrordir makes a minimal set of changes to the directory <mirror> to make it identical to the directory <control>, thus you can simply set it download only the files that have changed (saving time and bandwidth, etc.)

All the transfers are uncompressed -- I'm not zipping or otherwise archieving via tar, etc. Nightly traffic is normally small (a couple of Megs typical), as only the new/updated files are downloaded (mirrored). (Log files and MySQL db's are backup via another method, but that's a different story...)

Requirements for a Linux machine to handle the backup are minimal. Take a: Pentium I (anything P60 and up) or better machine; about 64Meg ram; a NIC; one or more 1 gig (or larger) hard drive(s) and add RedHat 6.x or higher and you've got yourself a machine that can handle this simple task at < .5 CPU load. ;) I use mirrordir on a 2nd Linux machine, to mirror the mirror, as a 2nd backup source. I run the backup-box headless (no monitor attached) and keyboardless, accessing it, when needed, via Webmin.

The (2nd) mirror gives an exact copy of our Web "tree", that I can access locally and use to edit, update or test scripts on (via running Apache, PHP, etc. on the 2nd Linux machine) before uploading the changes to the "real" site.

Mirrordir -- see: http://mirrordir.sourceforge.net/

Originally posted by drewnick
It should not. Did you copy/paste the code I entered here?

Drew


Hi, yes, i used exactly the same code.

Originally posted by technoart
Try: tar -pcvf /home/sites.tar /home/sites/*

I tried this one, and whilst running everything looks great and gives no errors. This produced a 380MB file named sites.tar, which I downloaded just fine.

But upon extraction using Winrar I get a "corrupt archive" message and it only goes as far as extracting /home/sites/home/web

I wonder what the problem could be? I have tried tar'ing these folders so many times, and the same thing always happens no matter what switches I use.