All,
I recently installed XFS 1.0.1 on an IBM x340 w/ a ServeRAID controller,
root FS is still ext2, created a 400 GB filesystem on an already
existing partition (ext2 type 83 partition). NFS mounted by 16 client
machines, all of which write to the FS on completion of a compute
process, creating a single 12.9 GB file. We know that the write worked
correctly on ext2 NFS-mounted filesystem, so it's not
application-related (at least in terms of it performing the write
correctly in terms of queing/parrallelism). There were no unusual
entries in /var/log/messages (other than mount requests - which were
successful) on the NFS-serving machine or the clients, yet the write
still failed.
I am using the rpms from release 1.0.1, and the 2.4.9-13 (RedHat and
XFS) kernel from testing :
acl-1.0.7-0.i386.rpm
acl-devel-1.0.7-0.i386.rpm
anaconda-7.1-7XFS.i386.rpm
anaconda-runtime-7.1-7XFS.i386.rpm
attr-1.0.3-0.i386.rpm
attr-devel-1.0.3-0.i386.rpm
devfsd-1.3.11-sgi.i386.rpm
dmapi-0.1.1-0.i386.rpm
dmapi-devel-0.1.1-0.i386.rpm
kernel-doc-2.4.9-13SGI_XFS_PR1.i386.rpm
kernel-headers-2.4.9-13SGI_XFS_PR1.i386.rpm
kernel-smp-2.4.9-13SGI_XFS_PR1.i686.rpm
kernel-source-2.4.9-13SGI_XFS_PR1.i386.rpm
tux-2.1.0-2.i386.rpm
xfsdump-1.0.9-0.i386.rpm
xfsprogs-1.2.8-0.i386.rpm
xfsprogs-devel-1.2.8-0.i386.rpm
Has anyone had experience with this, have any help, hint, suggestions?
FYI, unless it's absolutely life-or-death, I would prefer to use the
RedHat-release kernels. Any help is very appreciated, thanks in
advance.
Regards,
Derek R.
--
Junior Linux Geek
713-817-1197 (cell)
713-781-4000 x2267 (office)
"Linux users, fanatical. No way...
HEY! Get that MCSE up on the altar,
Tux must be appeased!"
|