Wed, 4 Sep 2002 17:57:41 +0200
* Ben Escoto <firstname.lastname@example.org> [03. Sep. 2002]:
> >>>>> "GZ" == Gregor Zattler <email@example.com>
> >>>>> wrote the following on Tue, 3 Sep 2002 13:10:54 +0200
GZ>> How do i achive the copying of the incremets? Would cp -a
GZ>> /backup/*2002.07.* /backbackup catch all necessary files?
> >> No, the backup files will be in several directories in
> >> /backup/rdiff-backup-data.
GZ>> ?? but "cp -a" also traverses directorys?
> Yes, but the directories names do not necessarily contain "2002.07.".
> An increment called
> would not be copied by your command.
A mistake. So cp -a ...*2002-07-* will do the trick.
GZ>> My rdiff-backup slowly gets to large. Say i can hold only one
GZ>> month of increments on disk. I can --remove-older-than
GZ>> increments to free Diskussion space. But before i do that i
GZ>> want to backup this increments to CDs. I want to do this every
GZ>> month. So in the unprobably but not impossible case i want to
GZ>> restore a file in a the status months ago i can somehow
GZ>> restore the rdiff-backup directory from CD using say tar or cpio
GZ>> and then restore this file as of 2002-03-27 using rdiff-backup.
> Ok, I get it. As I think you mentioned earlier, rdiff-backup only
> adds files and directories to rdiff-backup-data, it never modifies or
> deletes any (except for the 2 log files, and it modifies directories
> in the sense that it adds files to them).
> So if you want a list of files that needs backing up you can
> probably just use find to find the new files, as in "find
> ./rdiff-backup-data -ctime -3" will find files newer than 3 days old.
> Offhand I don't know how to use this to copy them over, but it's
> probably possible using cp and xargs, or some other shell utility.
> (Perhaps someone else knows.) Worst case you will have do a bit of
> shell scripting...
Ahh. using "find" is a good idea.
Thanx for your patientness.
"The future is here. It's just not widely distributed yet."
-- William Gibson