rdiff-backup-users
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [rdiff-backup-users] Hidden directories are ignored during increment


From: Cybertinus
Subject: Re: [rdiff-backup-users] Hidden directories are ignored during incremental backup
Date: Wed, 03 Jun 2009 11:19:12 +0200
User-agent: Thunderbird 2.0.0.21 (Windows/20090302)

Hi,

Maarten Bezemer wrote:
> Hi,
>
>> On 01/06/09 18:45, John covici wrote:
>>> I think this is a problem in the fact that * never includes anything
>>> beginning wth . character.  To do this, you need to back up a superior
>>> directory to the Mail directory and that will work -- I do it all the
>>> time.
>
> If I look at the scripts, it seems that he is already starting the
> backup from /home/username/mail without using the * in the command line
> (which actually wouldn't make sense anyway).
>
>
> On Mon, 1 Jun 2009, Marcel (Felix) Giannelia wrote:
>> A good way I've found to get the dot files without going up a
>> directory (sometimes you really don't want the extra directory level)
>> is the pattern "* .[^.]*", e.g.
>>
>> ls * .[^.]*
>>
>> will give you a list of all files and directories in the current dir,
>> including dot files but without "." and ".."  -- I learned the hard
>> way that ".*" is a very bad idea because it does match "." and ".." :)
>
> That would be a bad idea, since it would not include a file named
> "..something". I've had users create files and directories with
> important stuff using such names, so that they would show up on top of
> the directory listing and be easy to find. So be careful with that!

That is true, there isn't any * in my script. But that also doesn't seam needed. I just put in directories and then rdiff-backup just copies them recursively. So I wouldn't know where I would need a /* behind a directory. The only reason for specifying a /* would be to exclude certain files, but then you need to make it a bit more useful regular expression ofcourse ;). When I read the manpage of rdiff-backup I don't get the impression that adding /* is needed.
>
>
> Tijn, you talk about IMAP and maildir. Typically, ~/mail is used to
> store mbox files (containing lots of messages in one file), and
> ~/Maildir is used to store Maildir / Maildir++ format mailboxes (in
> which each email is a separate file).
> So, please check your system to make sure you're not accidentally trying
> to backup the wrong mail location.

OK, that could be what people do most of the time. But the IMAP server I'm using I have set up myself and is running on my own computer. And I found ~/mail the most logical directory for my maildir, so I used that :). I didn't know about this custom. I don't create any mbox files on my computer. I find a maildir better then mbox, because when a mbox file gets corrupted all my mail is lost. When a single file in the maildir gets corrupted I only loose 1 mail. Which is a comforting idea imo ;). This is also one of the reasons I don't like Microsoft Exchange ;).
>
> Furthermore, I just checked my own setup, and I don't see any hidden
> files go missing from the backup. I didn't need to do anything special
> to get it done this way.
As it turns out, so does my system :). I was just looking with soup in my eyes (which is a Dutch saying ;) )
>
> I also noted that you use the '--create-full-path' parameter when
> starting the backup. Using it that way may with several paths in
> multiple runs may not do exactly what's expected.
> Do you want your backups to look like:
> /backup/snapshots/hourly/home/cybertinus/mail/home/cybertinus/mail/something?

That is not the behaviour I see from rdiff-backup. I just tell it to backup /home/cybertinus/mail to /backup/snapshots/hourly/home/cybertinus/mail. rdiff-backup notices the first time that the directorytree doesn't go any further then /backup/snapshots/hourly and thus creates the missing sublevels. The next run it notices that the entier needed tree exists and doesn't recreate it. It just increments the backup in /backup/snapshots/hourly/home/cybertinus/mail. Which is something I like :). This is exactly the behaviour I want, and when I read the manpage I also got the impression that this was the behaviour.

>
>
> Something you could do to find out exactly what rdiff-backup is doing,
> is adding a -v5 parameter to the command line, and see what it does. In
> the rdiff-backup-data you can also find session statistics files, mirror
> metadata files with information about every backed-up file, and a full
> backup log to browse through.

that -v5 option is nice. Maybe I'm gonna use it with every backup, but then at a bit less verbose level. I'm gonna test it a bit. Thnx anyway for the tip :).
>
> Hope this helps..

Yeah, it was a usefull post. Thnx :)
>
> Regards,
> Maarten

Regards,
Cybertinus
>
>
> _______________________________________________
> rdiff-backup-users mailing list at address@hidden
> http://lists.nongnu.org/mailman/listinfo/rdiff-backup-users
> Wiki URL:
> http://rdiff-backup.solutionsfirst.com.au/index.php/RdiffBackupWiki




reply via email to

[Prev in Thread] Current Thread [Next in Thread]