> 

> "Marty Rosenberg":
>> The subject pretty much says it all.  I've mounted a bunch of
>> directories with the command: mount -t aufs -o
>> dirs=/mnt/farm1:/mnt/farm2:/mnt/farm3:/mnt/farm4:/mnt/farm5:/mnt/farm6:
>> /mnt/farm7 none /data_test after which, attempting to find a particular
>> file in /data_test gives strange errors like: find:
>> ./videos/tv/Seinfeld/Seinfeld Season 3/Seinfeld 3 - 21 The Keys.avi/(Ep
>> 10 of 13) Fakin' It.avi: No such file or directory
> 
> Please let me make sure first. - you have configured aufs correctly,
> including last lhash.patch problem. - you didn't get any warnings both of
> at compiling and loading the module. - all your branches are writable, and
> they are all xfs.
> 
> I will test xfs with huge number of files, please wait for a while. Also,
> I am afraid that I have to check 2.6.18-gentoo-r6 kernel.
> 
>
 
sorry for not including the info that I should have.  I've removed the 
lhash.patch problem, there are no problems with compiling or loading the 
module.  This time, I've decided to use the kconfig method, and I've recompiled 
the entire kernel.  I'm using aufs 20070312.  

Some more information that's probably useful is that there doesn't seem to be 
any specific correlation between the mount points and the errors.  The 
confusion is specifically it assuming that a file is another directory that 
exists someplace else inside of the union.  This other directory is sometimes 
on the same partition as the place where it's accidentialy mounted, and 
sometimes it isn't.  It looks like it's rather consistent as to which files it 
believes are directories, and what directories it's pulling into other 
locations.  Another clarification: "screenfuls of errors" = 406 of them.

>> PS: I'm really looking forward to being able to export a union over
>> NFS.  is there anything I can do to help out? I've got plenty of boxes
>> with various platforms/ architectures.
> 
> After this problem fixes, I will send you a patch which exports aufs.
> Please test it. And please try 64bit environment and SMP machines. Aufs
> needs to be tested on such environments, because I don't have such
> machines.
> 
> Junjiro Okajima
> 
> 

I have an AMD 64-X2 and two old dual proc pentium 2's.  Currently they're both 
running dragonfly bsd, but that can change.


-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV

Reply via email to