Alternate path issue

[subeeshpazhuvil]

we are facing some alternate path issue in EMC clarion disks,
25 disks are showing alternate links failed ,we are using PVlinks there is no powerpath installed on this.But one strange thing ioscan output is showing NO H/W ,diskinfo showing proper disks size and all.Even inq command also not giving any informatio,Can anyone suggest the solution for this,w
Class I H/W Path Driver S/W State H/W Type Description
===========================================================================
disk 268 0/4/0/0.98.116.19.0.0.1 sdisk NO_HW DEVICE DGC CX700WDR5
/dev/dsk/c24t0d1 /dev/rdsk/c24t0d1
genesis#diskinfo /dev/rdsk/c24t0d1
SCSI describe of /dev/rdsk/c24t0d1:
vendor: DGC
product id: CX700WDR5
type: direct access
size: 193986560 Kbytes
bytes per sector: 512

[James R. Ferguson]

Hi:

"No_HW" means that the device is no longer responding. If you have 25 devices with "NO_HW" you should look at a channel or controller failure or dis-connection. Your EMS should have logged information to aid you in diagnosis.

Regards!

...JRF...

[Sandman!]

The loss of 25 LUNs indicates losing a fibre channel path or an HBA. Check the output of the following commands...

# ioscan -funC fc
# ioscan -funC fcp

Verify if a previously CLAIMED fc / fcp path now shows NO_HW.

[subeeshpazhuvil]

25 disks means it's not from same path,other disks are visible through the same controller.Pls check my ioscan out put.

[Sandman!]

Could you attach output of the command below:

# ioscan -funC fcp

~thanks

[subeeshpazhuvil]

genesis#ioscan -funC fcp
Class I H/W Path Driver S/W State H/W Type Description
====================================================================
fcp 0 0/4/0/0.97 fcp CLAIMED INTERFACE FCP Domain
fcp 2 0/4/0/0.98 fcp CLAIMED INTERFACE FCP Domain
fcp 3 1/10/0/0.97 fcp CLAIMED INTERFACE FCP Domain
fcp 1 1/10/0/0.98 fcp CLAIMED INTERFACE FCP Domain

[John Guster]

check SAN storage team to see if it is zoned multiple patchs to this host.

[John Guster]

you see NO_HW from diskinfo becasue you have auto tresspass enabled. One of the path is dead at EMC side as your HBA seems working.

[Minoia]

Sometime happen with Clariion and HP-UX.
Please have a look at EMC2 manual "EMC Host Connectivity Guide for HP-UX" and verify carefully the connection parameter.
To solve your problem run on Navisphere Tools -> Failover Setup Wizard.
The wizard help you to define the right connection parameter at the end run ioscan on HP-UX server.

[Jollyjet]

use fcmsutil and check the connectivity is ok.

[John McWilliams_1]

Check the EMC powerlink website for the e-lab Interoperability Navigator. In there you can search for the CX model, HBA, and HPUX version and it will describe patches, and requirements. There are versions of HP-UX and CLARiiON that will not work with PV links, therefore powerpath is required.
I have this same issue and as the system is temporary I used 1 path.
Cheers John