I'm experiencing some troubles using an aufs volume
exported by NFS.
I think that the following behaviour isn't expected.
mkdir A
touch A/B
ln A/B B
i.e. we have a directory with a file and a hard link to it.
echo test >> A/B --> Works
echo test >> B --> Works too, but there is a strange delay
now,
rm A/B
echo test >> B
--> Still working, but it takes a very long delay now
--> Another step
rmdir A
--> Okay, it won't be remove becauce .nfs* file
--> But it has a side effect, it brokes the open file descriptor:
tail: A/B: No such file or directory
tail: no files remaining
I've tried diferent aufs and kernel versions. Now I'm working with kernel
2.6.22.9 and the last aufs version.
The setup for this test was:
mount -n -t aufs -o
noplink,nowarn_perm,br:/export/rws/servers/caporal=rw:/export/mainroot=ro aufs
/mnt
exportfs -o no_root_squash,rw,fsid=1 estrella:/mnt
mount -o rw,nfsvers=3 estrella:/mnt /root/testeaufs
Thanks
Bona
p.s.: Another way to test it is trying to use the mutt mail reader with a /tmp
via NFS over
AUFS (mutt does anything very similar to it creating locking files)
-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems? Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >> http://get.splunk.com/
