Re: [OT] Backup solutions - rdiff

[stavel]

Hi Alvin,
> "Alvin" == Alvin Oga <[EMAIL PROTECTED]> writes:

Alvin> yes.. rdiff sounds like a good thing to do..  - make a 1
Alvin> line change and only save that one line diff changes

Alvin> - but if one creates a 10MB.file and 100MB.file and
Alvin> 1GB.file...  - thhose 10MB files have to be saved daily ...
Alvin> ( guess no different thant he current daily incrementals

:-)

Even you would use amanda you don't save disk space :-)

rdiff-backup technology saves disk space and net load but not
cpu on servers  (python) :-)

Alvin>  - problem is we cannot reset the "special directory"
Alvin> to now also include the new 10MB.file ... and want to only
Alvin> save the new changes of the 10MB.file

I didn't understand it. Each new 10MB file is a change. And it is
authomaticaly stored in 'special directory' when do new backup.

Alvin>  - if we do recreate a new "special directory", than
Alvin> the previous set of diff changes are lost ?? since those
Alvin> diffs was applied to a different master special directory
Alvin> ??

What is recreate a new 'special directory'?

If you delete this spec dir, you can't  recreate  backup.
This is dangerous to delete enything in stored tree since
reverse increments upon on it :-(

Because of that I have partition, the backups are stored in,
read-only.

Alvin> sounds like a fun problem to solve ??

:-)
-- 
  Jan Stavěl 


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: [OT] Backup solutions - rdiff

[Alvin Oga]

hi ya savel

On 21 Mar 2003 [EMAIL PROTECTED] wrote:

> Hi Bob,
> > "Bob" == Bob Paige <[EMAIL PROTECTED]> writes:
> 
> Bob> Opinions?
> 
> I tried rdiff-backup.
> I uses rsync-like transfer method to minimize net load and diff-like store method to 
> save
> incrementals. Diff method is extended even for binary file (that is rdiff algorithm).


yes.. rdiff sounds like a good thing to do..
- make a 1 line change and only save that one line diff changes

- but if one creates a 10MB.file  and 100MB.file  and 1GB.file...
- thhose 10MB files have to be saved daily ... 
( guess no different thant he current daily incrementals

- problem is we cannot reset the "special directory"  to now also
include the new 10MB.file ... and want to only save the new
changes of the 10MB.file

- if we do recreate a new "special directory", than the previous
set of diff changes are lost ?? since those diffs was applied to
a different master special directory ??

and yes.. a good way to do backups across the net ... little/minimal data
traffic

sounds like a fun problem to solve ??

c ya
alvin



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED] 
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]