[Top][All Lists]

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

[rdiff-backup-users] Hard link issues (was: Incredibly slow i/o to NAS s

From: Robert Nichols
Subject: [rdiff-backup-users] Hard link issues (was: Incredibly slow i/o to NAS server)
Date: Tue, 14 Feb 2017 13:03:14 -0600
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:45.0) Gecko/20100101 Thunderbird/45.7.0

On 12/05/2016 03:26 PM, Robert Nichols wrote:
On 12/05/2016 09:28 AM, Joe Steele wrote:

I believe the harlink problems you describe are identified in this bug report 
(that I submitted in 2009):


Attached to the report are patches to fix the issue.  Some distros (which I think 
include Debian, Ubuntu, & Suse) have since been including the patches as part 
of their packaging.

Those patches look like they address the problems with checksums being misplaced 
or lost, but they don't appear to have anything to do with what I see as the 
greater problem of sets of hard links being broken up and inconsistencies between 
the hard link counts in the metadata and the mirror.  I finally got rdiff-backup 
to build (needed a patch for librsync >= 1.0.0), so I can do some testing later.

Well, after two months of use I can confirm that those patches _do_ appear to 
solve the problem of hard links being broken up or inconsistent.  I still don't 
understand how, but it works.  My hard link audit hasn't found any issues since 
I installed those patches.

Doing the hard links correctly does aggravate the problem of tiny "zero-diff" 
files being created, mainly whenever a file has a change in its hard link set, but also 
under some other, hard to pin down conditions.  In a ~500GB archive I found I had 
accumulated over a million (10^6) such files, using ~4.3GB of space.  I now have a new 
audit that gets rid of those.  I've really got to make another stab a learning Python so 
that I can fix those problems at the source.  So far, I've only succeeded in being 
devoured by a snake.

Bob Nichols     "NOSPAM" is really part of my email address.
                Do NOT delete it.

reply via email to

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