Mounting LUN from EMC VNX5300 via FC on Centos 6

Question

On my VMX5300, I have created LUNs, configured hosts etc. Mounting LUNs in my ESXi is very simple, ESXi host discovered new FC device and with a few clicks I was able to mount new FC storage. On Centos it's not quite simple as I thought it would be.

I have Fibre Channel card: Emulex Corporation Saturn-X: LightPulse Fibre Channel Host Adapter (rev 03).

System sees it, after configuration of LUN in VNX those lines appeared in my logs:

Jul 28 13:50:37 server kernel: lpfc 0000:05:00.0: 0:1305 Link Down Event x2 received Data: x2 x20 x80000 x0 x0
Jul 28 13:51:07 server kernel: rport-3:0-0: blocked FC remote port time out: removing target and saving binding
Jul 28 13:51:07 server kernel: lpfc 0000:05:00.0: 0:(0):0203 Devloss timeout on WWPN 50:06:01:69:3e:a0:63:57 NPort x0000ef Data: x0 x7 x0
Jul 28 13:57:23 server kernel: lpfc 0000:05:00.0: 0:1303 Link Up Event x3 received Data: x3 x1 x20 x2 x0 x0 0
Jul 28 13:57:23 server kernel: lpfc 0000:05:00.0: 0:1309 Link Up Event npiv not supported in loop topology
Jul 28 13:57:23 server kernel: lpfc 0000:05:00.0: 0:(0):2858 FLOGI failure Status:x3/x18 TMO:x0
Jul 28 13:57:23 server kernel: lpfc 0000:05:00.0: 0:(0):2858 FLOGI failure Status:x3/x18 TMO:x0
Jul 28 13:57:23 server kernel: lpfc 0000:05:00.0: 0:(0):2858 FLOGI failure Status:x3/x18 TMO:x0
Jul 28 13:57:23 server kernel: lpfc 0000:05:00.0: 0:(0):0100 FLOGI failure Status:x3/x18 TMO:x0
Jul 28 13:57:23 server kernel: scsi 3:0:0:0: Direct-Access     DGC      LUNZ             0531 PQ: 0 ANSI: 4
Jul 28 13:57:23 server kernel: sd 3:0:0:0: Attached scsi generic sg5 type 0
Jul 28 13:57:23 server kernel: sd 3:0:0:0: [sdd] READ CAPACITY failed
Jul 28 13:57:23 server kernel: sd 3:0:0:0: [sdd] Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE
Jul 28 13:57:23 server kernel: sd 3:0:0:0: [sdd] Sense Key : Illegal Request [current] 
Jul 28 13:57:23 server kernel: sd 3:0:0:0: [sdd] Add. Sense: Logical unit not supported
Jul 28 13:57:23 server kernel: sd 3:0:0:0: [sdd] Test WP failed, assume Write Enabled
Jul 28 13:57:23 server kernel: sd 3:0:0:0: [sdd] Asking for cache data failed
Jul 28 13:57:23 server kernel: sd 3:0:0:0: [sdd] Assuming drive cache: write through
Jul 28 13:57:23 server kernel: sd 3:0:0:0: [sdd] READ CAPACITY failed
Jul 28 13:57:23 server kernel: sd 3:0:0:0: [sdd] Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE
Jul 28 13:57:23 server kernel: sd 3:0:0:0: [sdd] Sense Key : Illegal Request [current] 
Jul 28 13:57:23 server kernel: sd 3:0:0:0: [sdd] Add. Sense: Logical unit not supported
Jul 28 13:57:23 server kernel: sd 3:0:0:0: [sdd] Test WP failed, assume Write Enabled
Jul 28 13:57:23 server kernel: sd 3:0:0:0: [sdd] Asking for cache data failed
Jul 28 13:57:23 server kernel: sd 3:0:0:0: [sdd] Assuming drive cache: write through
Jul 28 13:57:23 server kernel: sd 3:0:0:0: [sdd] Attached SCSI disk

So I assume that system discovered something :) ?

In /dev directory there is a new /dev/sdd device but running fdisk /dev/sdd gives me Can not read /dev/sdd.

From what I read I think that for mounting I should use multipath (/dev/mapper) but have no idea how to do that.

But fdisk -l doesn't show my new storage. I was googling but didn't find any simple (simple for me and my skills) answer or tutorial on how to do that...

Solution 1

I tried the Chopper3 answer but that didn't help. If block device appeared in /dev fdisk I should see it and after making partition and filesystem I should mount it without problems.

I couldn't restart the machine because it was in production use. So I unloaded my FC card drivers from kernel in my case

modprobe -r lpfc

and loaded the driver once again with

modprobe lpfc

After that a new block device appeared but this time everything was ok and fdisk was able to see the new device. Partitioning, formating to ext4 and now I have new FC LUN mounted in my system

:)

Solution 2

From what you'd copied into your question it looks like either a very simple problem or a much more complex mispresentation or potentially a zoning issue, but first let's do the easy option by trying rescan using;

echo “- – -” > /sys/class/scsi_host/host#/scan

where '#'is probably the number zero but you can check this by traversing into that directory.

Then do the 'fdisk -l' again, see if anything's changed and let us know.

Comments

[B14D3]

I have there five directories host0 host1 host2 host3 host4 ??

[Chopper3]

No real harm doing this to them all - it pauses/can-pause them while it does it though, just for a second or two - just want you to be aware.