Hello everybody!

I am building a failover storage system with two SunFire X2100 M2 server 
systems and a external RAID array for data. The array is connected via fibre 
channel.

I'm using Solaris Express Community Edition snv_86 with the supported Solaris 
Cluster Express Package. The installation went fine. The fibre channel array 
has one LUN per disk and the LUNs are building a zpool.

Now I tried to get NFS high available. I followed the NFS cluster data service 
documentation on docs.sun.com. But when I try to share a directory on the Zpool 
I get an system error. But the logs are quiet. I also checked the blog entry, 
to use not the sharenfs option. 
http://blogs.sun.com/SC/entry/why_sunw_nfs_is_required

Okay. Let's make it visual ;-)

tank is my ZFS pool. The pool ist managed by HAStoragePlus. On tank I have the 
"nfs" directory for the HA-NFS information. It's (hopefully) just like 
described in the documentation.

I also have a directory called "data" which I want to share for testing 
purposes.

But when I try to share the directory I get the following message

root at storage1:/# share -F nfs -o rw /tank/data
Could not share: /tank/data: system error

I also tried to make a ZFS for data, but I get the same error. The thing is, 
that I can share local directories like /mnt or something and it works fine. 
It's just when I try to share a directory on the HAStoragePlus ZFS pool.

Here is my installation history

zpool create tank  /dev/did/dsk/d1 /dev/did/dsk/d2 /dev/did/dsk/d3 
/dev/did/dsk/d4 /dev/did/dsk/d5 /dev/did/dsk/d6 /dev/did/dsk/d7
clresourcetype register SUNW.HAStoragePlus SUNW.nfs
mkdir /tank/nfs
clresourcegroup create -p PathPrefix=/tank/nfs storage    
clreslogicalhostname create -g storage -h storage VirtualIP
clresource create -g storage -t SUNW.HAStoragePlus -p Zpools=tank tank
clresourcegroup online -M storage
mkdir /tank/nfs/SUNW.nfs
mkdir /tank/nfs/export
mkdir /tank/data
cp /etc/dfs/dfstab /tank/nfs/SUNW.nfs/dfstab.nfs
vi /tank/nfs/SUNW.nfs/dfstab.nfs

# Do not modify this file directly.
# Use the sharemgr(1m) command for all share management
# This file is reconstructed and only maintained for backward
# compatibility. Configuration lines could be lost.
#
#       share [-F fstype] [ -o options] [-d "<text>"] <pathname> [resource]
#       .e.g,
#       share  -F nfs  -o rw=engineering  -d "home dirs"  /export/home2
share   -F nfs  -o rw   /tank/data

clresource create -g storage -t SUNW.nfs -p Resource_dependencies=tank nfs
clresourcegroup online -M storage

The resource group goes offline because of the NFS error.

I have the strange feeling that I'm missing something but my experience with 
the cluster framework is too little that I can figure out by my own.

I hope someone has an idea on this one. Thanks in advanced. I hope I haven't 
missed any details. :-)

Best regards
Chris
--

This message posted from opensolaris.org


Reply via email to