External Storage Provision (vnx) on Itanium ia64 hp Integrity rx2800 i4 running HP11.31

[Mark Parsons]

Hi, I am trying to add vnx storage of a 40gb lun to my Itanium server ia64 hp Integrity rx2800 i4 running HP11.31. However it is appearing as either 0gb or 2x40gb and it should show 4 paths and not 2. Please see screenshot attached document. Any suggestions? Many thanks in advance.

[Bill Hassell]

The two paths are unique LUNs so the VNX setup is providing two different LUNs, not one lun with 4 paths. SMH appears to be showing them with 2 paths to each unique LUN. I would try using the command line tools rather than SMH and taking snapshots of screens. Unix systems are based on command line interfaces and GUIs like SMH tend to obscure the details and taking a picture makes copying the text impossible. You'll want to be using Notepad and copying the all text to your posts.

To scan for new disk LUNs:

# ioscan -nfCdisk

If you are running 11.31:

# ioscan -m dsf

[Robert Herron]

How do you have the host configured in Unisphere > ARRAY > Hosts > Host List, Details pane, Initiators tab?  I have a rx6600, a rx2600, and a rx2800 i2 all running HP-UX 11.31 using storage from a VNX 5500.  Each host's initiators are configured for "Initator Type = CLARiion/VNX" and "Failover Mode = Active/Active mode (ALUA) - failovermode 4".  My hosts see the disk properly with this configuration.

What is the "World Wide ID" if you drill down into each of the 40GB LUNS?  If they have different IDs, then it would seen there are two different LUNs presented to the rx2800i4.  (The "World Wide ID" in SMH is the "Unique ID" show in Unisphere > ARRAY > Storage > LUNs > properties and more commonly known as the WWN.

[Mark Parsons]

We have checked and all initiators now are configured for "Initator Type = CLARiion/VNX" and "Failover Mode = Active/Active mode (ALUA) - failovermode 4".  

Each lun also has a different wwn / wwi.

I've attached the latest screenshot (san2.doc). Interestingly one of the 40gb has now disappeared. I am also wondering where this LUNZ is appearing from and not VNX5700WDVR.

[Robert Herron]

It is normal to see the "DGC LUNZ" is normal with VNX connected to the a HP-UX host.  In simple terms, LUNZ represents the VNX array and allows a sysadmin to verify the array is visible.  All my HP-UX hosts connected to VNX's list zero paths in SMH for LUNZ.

So, my guess is your SAN admin removed the now-missing LUN from your rx2800i4 host's storage group.  Check your syslog for any recent entries containing "scsimgr replace_wwid". 

Also, have your SAN admin verify that all your FC-HBA are zoned correctly in your FC switches and added to the VNX.  Run a "ioscan -fnNC fc" to list the device file for each FC-HBA.  Then run a 'fcmsutil <devfile> |grep "N_Port Port World Wide Name" '

EDIT:  Finishing my incomplete thought:

...fcmsutil....  The "N_Port WWN" for each FC-HBA are the WWNs used when creating the zones to the host and array to see each other.

[Mark Parsons]

Thanks for that - the storage guy in my firm that I deal with has now gone on vacation for 12 days. The issue isnt important (yet!) as the servers are not in production at the moment. I will be back in touch as soon as he returns.

[Torsten.]

If the LUN was unpresented and presented again, the server will not access the LUNs unless you do some scsimgr commands - or - reboot the server. Did you try this?

[Mark Parsons]

Yes - a reboot was performed.