Is it possible to present multiple LUNs from 1 SAN with 1 NFS server? Currently, my NFS server is directly connected to my FC SAN. Do I have to install and configure multipathing for this to work OR is this just not possible to do over one connection?

You can present as many LUNs from a SAN to one HBA as you want.

Well, I am currently having a problem where I am adding an additional LUN on my SAN BUT my Ubuntu 10.04 server doesn't allow me to fdisk the new LUN. I can see that it is defintely connecting to the server BUT I get an error when I try to run fdisk:

$ sudo fdisk /dev/sdb

Unable to read /dev/sdb

I can see the new lun in /proc/scsi/scsi. I was thinking it was because I can't do multiple luns to one HBA, also wasn't sure that it was a multi-pathing issue since I just have one physical path...Do you know what command I am missing to get this going?

Thanks

You only need one path. You don't need multipath when you only have one path.

Try where 0 is the right host (or all of them)

Or have you tried rebooting?

Ok, so I have 5 hosts? I tried that command with all 5 hosts and I didn't get any info back from it.

Yes, I have tried rebooting the server a few times but that doesn't help either. Very strange that I can run many commands that give me information regarding the storage but I cannot use it. Here is info on both LUNs that are connected (disk:0 is working and was setup prior to this new one disk:1):

sudo lshw -C disk

*-disk:0
description: SCSI Disk
product: 1726-4xx FAStT
vendor: IBM
physical id: 0.0.1
bus info: scsi@4:0.0.1
logical name: /dev/sda
version: 0617
size: 12TiB (13TB)
capacity: 12TiB (13TB)
capabilities: 7200rpm partitioned partitioned:dos
configuration: ansiversion=5

-disk:1
description: SCSI Disk
product: 1726-4xx FAStT
vendor: IBM
physical id: 0.0.2
bus info: scsi@4:0.0.2
logical name: /dev/sdb
version: 0617
size: 10TiB (11TB)
capacity: 10TiB (11TB)
capabilities: 7200rpm
configuration: ansiversion=5

Thanks in advance for any other ideas..

Another clue I have found, when I am directly logged into the box, I see a error like the one below repeating around every 5 secs or so:

"[65026.040045] end_request: I/O error, dev sdb, sector 0"

as well as this in most of my logs:

[66327.900412] sd 4:0:0:2: [sdb] Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE
Apr 28 10:17:07 nfs1 kernel: [66327.900419] sd 4:0:0:2: [sdb] Sense Key : Illegal Request [current]
Apr 28 10:17:07 nfs1 kernel: [66327.900425] sd 4:0:0:2: [sdb] <<vendor>> ASC=0x94 ASCQ=0x1ASC=0x94 ASCQ=0x1
Apr 28 10:17:07 nfs1 kernel: [66327.900434] sd 4:0:0:2: [sdb] CDB: Read(10): 28 00 00 00 00 00 00 00 08 00
Apr 28 10:17:07 nfs1 kernel: [66327.900446] end_request: I/O error, dev sdb, sector 0

Found the solution to this. Wasn't an OS problem. The problem was on the SAN that I was using. My SAN has 2 controllers but I am currently only using controller A. So, when I created the new array and logical drive, the path was being defaulted to controller B, even though it wasn't plugged in. The LUN information got passed to the OS through controller A, which is what I was seeing BUT the path to the LUN wasn't available.