duplicity-talk
[Top][All Lists]
Advanced

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

Re: [Duplicity-talk] Continuing an interrupted backup.


From: Peter Schuller
Subject: Re: [Duplicity-talk] Continuing an interrupted backup.
Date: Fri, 2 Nov 2007 19:18:20 +0100
User-agent: Mutt/1.5.16 (2007-06-09)

> > (1) Possibly implementing expoential backoff for re-tries to make it a
> > bit more general, but in particular
> 
> This would have to be done for all the backends.  Not a real problem,
> just a note.

Code duplication alert went off in my head here ;) I'll look into
re-designing things a bit so that retry happens at a level further
up. I don't see any kink here based on what I've seen in the backend
code so far, but will have to look into it properly.

> Simpler would be a file like '/path/duplicity-halted' which could be
> removed in order to restart.  You could also put the error reason in
> the file itself.  Or, if you do use a socket, you could actually post
> progress messages to the socket... just a thought.

I think I like the create-file-with-error-reason approach best,
although technically a socket can be infinitely more flexible.

> This could be handy.  Command line is getting long!

Long indeed.

I have been thinking of writing a higher level tool to handle backups,
that would have backends for duplicity, rdiff-backup, and perhaps
others. Exactly because with all concerns (history retentation,
frequency of full vs. incremental, unattended-yet-error-resilient,
etc) you end up with long command lines in various combinations. As a
result I end up with wrapper scripts that are semi-general, but
specific enough that I can't use the same script in different
situations.

An alternative would be to implement some higher level features in
duplicity itself, but I think I prefer the former approach becauas the
payoff is higher (works with multiple backup backends), and it doesn't
imply bloat in the core tool.

If anyone has opinions on this, feel free to chime in. (Not promising
I'll actually do this; just something I'm thinking of.)

-- 
/ Peter Schuller

PGP userID: 0xE9758B7D or 'Peter Schuller <address@hidden>'
Key retrieval: Send an E-Mail to address@hidden
E-Mail: address@hidden Web: http://www.scode.org

Attachment: pgpx_NIkYorn2.pgp
Description: PGP signature


reply via email to

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