[Gluster-users] nfs/server not loading
Shehjar Tikoo
shehjart at gluster.com
Fri Aug 20 05:45:03 UTC 2010
Jesse Caldwell wrote:
> hi all,
>
> i just built glusterfs-nfs_beta_rc10 on freebsd 8.1. i configured
> glusterfs as follows:
>
> ./configure --disable-fuse-client --prefix=/usr/local/glusterfs
The volume file looks fine. We've never tried anything with the beta
branch on fbsd. Let me see if I can get it setup for a few build tests
atleast. In the mean time, please let me have the complete log file from
the glusterfsd that runs nfs/server. Use the TRACE log level by setting
the following command line options:
-L TRACE -l /tmp/nfs-fail.log
Mail me the nfs-fail.log.
Thanks
>
> i also ran this on the source tree before building:
>
> for file in $(find . -type f -exec grep -l EBADFD {} \;); do
> sed -i -e 's/EBADFD/EBADF/g' ${file};
> done
>
> i used glusterfs-volgen to create some config files:
>
> glusterfs-volgen -n export --raid 1 --nfs 10.0.0.10:/pool 10.0.0.20:/pool
>
> glusterfsd will start up with 10.0.0.10-export-export.vol or
> 10.0.0.20-export-export.vol without any complaints. when i try to start
> the nfs server, i get:
>
> nfs1:~ $ sudo /usr/local/glusterfs/sbin/glusterfsd -f ./export-tcp.vol
> Volume 'nfsxlator', line 31: type 'nfs/server' is not valid or not found on this machine
> error in parsing volume file ./export-tcp.vol
> exiting
>
> the module is present, though, and truss shows that glusterfsd is finding
> and opening it:
>
> open("/usr/local/glusterfs/lib/glusterfs/nfs_beta_rc10/xlator/nfs/server.so",O_RDONLY,0106) = 7 (0x7)
>
> nfs/server.so doesn't seem to be tragically mangled:
>
> nfs1:~ $ ldd /usr/local/glusterfs/lib/glusterfs/nfs_beta_rc10/xlator/nfs/server.so
> /usr/local/glusterfs/lib/glusterfs/nfs_beta_rc10/xlator/nfs/server.so:
> libglrpcsvc.so.0 => /usr/local/glusterfs/lib/libglrpcsvc.so.0 (0x800c00000)
> libglusterfs.so.0 => /usr/local/glusterfs/lib/libglusterfs.so.0 (0x800d17000)
> libthr.so.3 => /lib/libthr.so.3 (0x800e6a000)
> libc.so.7 => /lib/libc.so.7 (0x800647000)
>
> is this a freebsd-ism, or did i screw up something obvious? the config
> file i am using is obviously nothing special, but here it is:
>
> nfs1:~ $ grep -v '^#' export-tcp.vol
>
> volume 10.0.0.20-1
> type protocol/client
> option transport-type tcp
> option remote-host 10.0.0.20
> option transport.socket.nodelay on
> option transport.remote-port 6996
> option remote-subvolume brick1
> end-volume
>
> volume 10.0.0.10-1
> type protocol/client
> option transport-type tcp
> option remote-host 10.0.0.10
> option transport.socket.nodelay on
> option transport.remote-port 6996
> option remote-subvolume brick1
> end-volume
>
> volume mirror-0
> type cluster/replicate
> subvolumes 10.0.0.10-1 10.0.0.20-1
> end-volume
>
> volume nfsxlator
> type nfs/server
> subvolumes mirror-0
> option rpc-auth.addr.mirror-0.allow *
> end-volume
>
> thanks,
>
> jesse
> _______________________________________________
> Gluster-users mailing list
> Gluster-users at gluster.org
> http://gluster.org/cgi-bin/mailman/listinfo/gluster-users
More information about the Gluster-users
mailing list