Clariion CX700 disk on HPUX 11.23

[maslo64]

Hi all,
I am facing weird issue with CLARiiON CX700.
HPUX 11.23
EMCpower HP.4.5.0_b069 PowerPath IA64
NAVIAGENT 6.24.6.0.32 Navisphere Disk Array Management Tool (AGENT)

I Have two nodes where are the disks mapped and I can access them from node1 and created VolumeGroup, LV and Filesystem.
But when I tried do pvdisplay from node2 it fail with IO Error.

disk 570 0/3/1/0.225.14.0.0.0.0 sdisk CLAIMED DEVICE DGC CX700WDR5
disk 569 0/4/1/1.227.16.0.0.0.0 sdisk CLAIMED DEVICE DGC CX700WDR5

node1:
(root)#powermt display class=clariion
CLARiiON logical device count=1
=========================================================
----- Host Bus Adapters --------- ------ I/O Paths ----- ------ Stats ------
### HW Path Summary Total Dead IO/Sec Q-IOs Errors
=========================================================
92 0/3/1/0.225.14.0.0 optimal 1 0 - 0 0
93 0/4/1/1.227.16.0.0 optimal 1 0 - 0 0


node2:
(root)#powermt display class=clariion
CLARiiON logical device count=0
=========================================================
----- Host Bus Adapters --------- ------ I/O Paths ----- ------ Stats ------
### HW Path Summary Total Dead IO/Sec Q-IOs Errors
=========================================================

NODE2:(root)#ioscan -fnH0/3/1/1.227.16.0.0.0.0
Class I H/W Path Driver S/W State H/W Type Description
=========================================================
disk 30 0/3/1/1.227.16.0.0.0.0 sdisk CLAIMED DEVICE DGC CX700WDUNB
/dev/dsk/c86t0d0 /dev/rdsk/c86t0d0
NODE2:(root)#pvdisplay -l /dev/dsk/c86t0d0
pvdisplay: IO error on Physical Volume device "/dev/dsk/c86t0d0".
NODE2:(root)#


Thanks for any direction.
Ladislav

[Jeeshan]

check the host integration is properly performed or not?

if you want to see same LUN from both nodes than check the both hosts are members of same storage group.

also check the following command

#diskinfo /dev/rdsk/c86t0d0

[maslo64]

Hi,
Our storage guys confirmed that LUNs are presented OK for both nodes.

NODE1:(root)#diskinfo /dev/rdsk/c92t0d0
SCSI describe of /dev/rdsk/c92t0d0:
vendor: DGC
product id: CX700WDR5
type: direct access
size: 209715200 Kbytes
bytes per sector: 512
NODE1:(root)#

NODE2:(root)#diskinfo /dev/rdsk/c86t0d0
SCSI describe of /dev/rdsk/c86t0d0:
vendor: DGC
product id: CX700WDUNB
type: direct access
size: 0 Kbytes
bytes per sector: 0
NODE2:(root)#

[Steven E. Protter]

Shalom,

It would appear the World Wide Name WWN of the node2 is not configured into the disk array.

That is the first thing I'd check.

SEP

[maslo64]

Hello SEP,
if there is not configured correct WWN on Array/SAN ,it will not appear in system at all. Am I right ? But device files are visible from system.
Only difference which I see is in ioscan.
NODE1 : "CX700WDR5"
NODE2 : "CX700WDUNB"

[smatador]

Hi,
For me powerpath on node2 is not correct. You don't see anything of the clariion. So after that HPUX could'nt see the pv
Could you do a
# powermt check
# powermt config
if is ok do a powermt save

[maslo64]

Hi,
This was first think I did after LUNs were presented to the system.
Anyway this didn`t help. I can see disks from DMX array , but CLARiiON shows nothing.

NODE2:(root)#powermt check
NODE2:(root)#powermt config
NODE2:(root)#powermt display class=clariion
CLARiiON logical device count=0
======================================================================
----- Host Bus Adapters --------- ------ I/O Paths ----- ------ Stats ------
### HW Path Summary Total Dead IO/Sec Q-IOs Errors
======================================================================


NODE2:(root)#
NDOE2:(root)#powermt display
Symmetrix logical device count=9
CLARiiON logical device count=0
....

[smatador]

Hi,
Maybe you have to check the entire procedure starting with what is the FLARE supported on your CX7000
EMC CLARiiON CX700/CX500 HPQ HP-UX 11i v2 (HP-UX 11.23)1 EMC PowerPath: 4.5.0, 4.5.1, 5.0.0, 5.0.1, 5.1 EMC FLARE 02.16, EMC FLARE 02.17, EMC FLARE 02.19, EMC FLARE 02.24, EMC FLARE 02.26
If you have access to powerlink.emc.com? You could check elab interoperability and have more information

[maslo64]

Hi,
Checking if there is not problem described in "ID: emc182688" @ powerlink

Problem: Hosts not logging in to CLARiiON array on all zoned paths.

Thanks for point.

[D Anderton]

Verify they are in the same storage group in Navisphere.

Also that the fabric is logged in.

Go onto EMC powerlink and verify the initiator settings + host connectivity guide.

[Jeeshan]

>>disk 30 0/3/1/1.227.16.0.0.0.0 sdisk CLAIMED DEVICE DGC CX700WDUNB

is not LUN. its a fake zoning device which indicates you have 1 HBA zoning.


Did you apply command the following commands after zoning?

#insf (installing driver of assigned LUN)

#powermt config
#powermt save
#powermt display dev=all

(you need to apply the above commands serially and no break.)

[maslo64]

Hi,
Yes, I did use these command extactly in that order.

[Jeeshan]

now, whats the status?

if the problem still exists then check the host integration or zoning or you can stop and start the naviagent.

[maslo64]

Status did not change. Disk are still not accessible.our storage guys confirm that LUN is presented correctly. Also I tried to restart the agent with no effect. Now waiting on they response to note "emc182688"

[dinil]

HI,

I would suggest the following steps.
1. check if the Initiators of node 2 are properly logged in and registered into the clariion box. This is seen from the navisphere connectivity status.
2. See if both the hosts are in same storage group. Also from the navisphere manager, enable the engineering mode, where in you will get the "connect hosts" session on clicking the storage group, where you can see the "Advanced" tab. You should enable all the four ports of the clariion (A0,A1,B0 andB1) for both nodes depending on the zoning used.
3. do a ioscan -fnC disk from the node 2
4. insf
5.powermt config
6.powermt display class=clariion

Check the powermt display options to see if clariion devices are managed/unmanaged.

[maslo64]

Hi,
I have now confirmed that we are not affected by "emc182688".


>1. check if the Initiators of node 2 are properly logged in and registered into the clariion box. This is seen from the navisphere connectivity status.

This was good point NODE1 is Logged IN and Registered. But NODE2 is Registered , but it`s NOT Logged In.

And now the question. How I can register it ?

Thanks

[Steven E. Protter · Accepted solution]

Shalom again,

I think the SAN might be improperly zoned.

If so, it is possible systems are seeing disks that are not assigned to them via WWN

SEP

[TTr]

> disk 30 0/3/1/1.227.16.0.0.0.0 sdisk CLAIMED DEVICE DGC CX700WDUNB

The disk 30 in node2 appears to be the same as disk 569 in node1 based on the FC domain ID and array port number "...227.16..." in the hardware paths. If they are indeed the same disk, I am pointing out that in node1 the disk identification string in ioscan shows up as "CX700WDR5" which is a valid device, whereas in node2 the same string is "CX700WDUNB". There is something wrong here "UNB" stands for UNBound. I would go back to the disk array admin and ask for details in the presentation of the disks. Or the two disks in node1 and node2 might not be the same LUN.

[maslo64]

Hi all,
Thanks you all who gives me advices.
Suddenly it working "somehow" after kicking to storage guys.

[maslo64]

Mostlikely problem was in Navisphere/SAN somwhere. It`s not clear to me or storage guys.