linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* Data Corruption with Dump
@ 2002-05-20 11:40 Ganajyoti Bhattacherjee
  2002-05-20 12:03 ` Michael Sinz
  0 siblings, 1 reply; 2+ messages in thread
From: Ganajyoti Bhattacherjee @ 2002-05-20 11:40 UTC (permalink / raw)
  To: linux-kernel


I just happenned to read Linus'opinion on using dump for backup (in a
mail
from him to linux-kernel mailing list on 27 Apr 2001). He wrote

-------------------------------------------------------------------------
"...Note that dump simply won't work reliably at all even in 2.4.x: the
buffer cache and the page cache (where all the actual data is) are not
coherent. This is only going to get even worse in 2.5.x, when the
directories are moved into the page cache as well.

So anybody who depends on "dump" getting backups right is already playing
russian rulette with their backups.  It's not at all guaranteed to get the
right results - you may end up having stale data in the buffer cache that
ends up being "backed up"......"
---------------------------------------------------------------------------

But is the above true even in case of taking "dump" of an unmounted
filesystem.
I am using SuSE 7.3 (kernel-2.4.18-SMP).

Thanx

--Gana
(G. Bhattacherjee)



^ permalink raw reply	[flat|nested] 2+ messages in thread

* Re: Data Corruption with Dump
  2002-05-20 11:40 Data Corruption with Dump Ganajyoti Bhattacherjee
@ 2002-05-20 12:03 ` Michael Sinz
  0 siblings, 0 replies; 2+ messages in thread
From: Michael Sinz @ 2002-05-20 12:03 UTC (permalink / raw)
  To: Ganajyoti Bhattacherjee; +Cc: linux-kernel

Ganajyoti Bhattacherjee wrote:
> I just happenned to read Linus'opinion on using dump for backup (in a
> mail from him to linux-kernel mailing list on 27 Apr 2001). He wrote
> 
> -------------------------------------------------------------------------
> "...Note that dump simply won't work reliably at all even in 2.4.x: the
> buffer cache and the page cache (where all the actual data is) are not
> coherent. This is only going to get even worse in 2.5.x, when the
> directories are moved into the page cache as well.
> 
> So anybody who depends on "dump" getting backups right is already playing
> russian rulette with their backups.  It's not at all guaranteed to get the
> right results - you may end up having stale data in the buffer cache that
> ends up being "backed up"......"
> ---------------------------------------------------------------------------
> 
> But is the above true even in case of taking "dump" of an unmounted
> filesystem.

No - this was written about mounted filesystems.  Since dump does not go
through the VFS layer and thus does not go through the buffer cache, it
will miss things in an active filesystem.  In a filesystem that is not
mounted (I assume cleanly unmounted) there would be no problem.  To me,
the issue of dump is that it really does not work on a always-up system.
I can not just take the system off-line to backup so the backup mechanisms
I use are all via the file system layer and thus work with the VFS and
page cache and buffer cache rather than around them...

-- 
Michael Sinz ---- Worldgate Communications ---- msinz@wgate.com
A master's secrets are only as good as
	the master's ability to explain them to others.



^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2002-05-20 12:03 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2002-05-20 11:40 Data Corruption with Dump Ganajyoti Bhattacherjee
2002-05-20 12:03 ` Michael Sinz

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).