Re: converting decremental to full backup

Matus UHLAR - fantomas <[email protected]>
Newsgroups gmane.comp.sysutils.backup.dar.support
Message-ID <[email protected]>
>On 01/06/2019 21:09, Matus UHLAR - fantomas wrote:
>> I have created a decremental backup and a full one, from older
>> time.
>>
>> Now I'd like to get rid of the full backup and convert the
>> decremental to full one. - in fact the opposite what I did when I
>> created the decremental backup.
>>
>> I have tried:
>>
>> dar -+  -A full -@ decremental -/ Oo # to try overwriting deta in
>> full archive with data from decremental archive
>>
>> I have tried replacing -A and -@, change -Oo to -Pp
>>
>> but either I get incremental/decremental backup, or I get the full
>> backup containing new files.
>>
>> is it possible to create old full debug without extracting both
>> backups?

On 02.06.19 12:03, Denis Corbin wrote:
>yes it is:
>
>assuming:
>- - at time 1 you have created a full backup named full1
>- - at time 2 you have created a second full backup named full2
>- - then just after you created a decremental backup decr1 which will
>replaces full1 (dar -+ decr1 -A full1 -@full2 -ad ...)
>- - you then deleted full1 as you now have decr1

exactly like that, I went through the docs and examples ...

>Your question is: how to recreate full1 using full2 and decr1, right?
>
>Just use this:
>
>dar -+ressurected_full1 -A full2 -@ decr1 full-from-diff
>
>full-from-diff target is defined in /etc/darrc and should be fetched
>by  by dar unless you add -N option.
>
>- ---
>full-from-diff:
>- -/ '{!(~I)}[Rr] {~S}[O*] P* ; {~s}[*o] *p'
>- ---

... but seems I didn't do through the docs deeply enough ...

>For the details, this overwriting is applied when a entry is found on
>both archives and means:
>- - "{!(~I)}" = if the entry in decr1 is not an inode, then "[Rr]" = do
>not add any entry in the resulting archive: The entry in decr1 is a
>"detruit" object that flag the file to be removed when restoring
>decr1, this is a file that was added at time 2.
>- - else "{~S}" = if the entry in decr1 is a fully saved one, "[O*]" =
>take it in the resulting archive else "P*" = use the entry from full2:
>file  has not changed between time 1 and 2, it is thus completely
>saved in full2 as it is a full backup
>- - for EA and FSA: "{~s}" = if the EA/FSA is fully saved in decr1 take
>it in the resulting archive else take the on from full2 which is
>always fully saved (full backup).

... because they seemed too cryptic for me.
but now, I hope it's documented and can be found in the net.

>last in case there is no conflict the overwriting policy does not
>apply, this should only occur if a file has been removed from time 1
>to time 2 which means a entry fully saved and only present in decr1,
>such entry is then added to the resulting archive, as expected.
>
>The situation where an entry is only present in full2 should not
>arrive: such entry would have been created at time 2 and doing the
>decremental backup, a "detruit" entry would have been added into decr1
>to indicate this entry to be removed from filesystem when restoring
>decr1 over full2, thus the overwriting policy would be triggered as
>explained above leading to completely ignore such entry.
>
>I hope the explanation are not too confusing :-)

I don't think so.
Thank you!
-- 
Matus UHLAR - fantomas, [email protected] ; http://www.fantomas.sk/
Warning: I wish NOT to receive e-mail advertising to this address.
Varovanie: na tuto adresu chcem NEDOSTAVAT akukolvek reklamnu postu.
I don't have lysdexia. The Dog wouldn't allow that.
lmpx.com only provides a reader for public news (NNTP) servers. It is not affiliated with the servers or forums shown here and is not responsible for the content of articles, which is written by their respective authors.