Host system fails to see Raid subsystem

From: Seela Balkissoon (seela@cs.yorku.ca)
Date: Wed Oct 23 2002 - 16:24:57 EDT


Hi managers,

   I am experiencing a failure on a system running an application that
   massages data and appends to a file. The file size has the potential
   to grow to 4GB.
    
   After several days of processing, the host system loses sight of the
   raid disks with the following error.
   
[snip]
 scsi: [ID 243001 kern.warning] WARNING: /pci@1f,2000/SUNW,ifp@1/ssd@w200000501
3b338d1,0 (ssd1):
   SCSI transport failed: reason 'timeout': retrying command
   scsi: [ID 243001 kern.warning] WARNING: /pci@1f,2000/SUNW,ifp@1/ssd@w2000005
013b338d1,0 (ssd1):
   SCSI transport failed: reason 'aborted': retrying command
   scsi: [ID 243001 kern.warning] WARNING: /pci@1f,2000/SUNW,ifp@1/ssd@w2000005
013b338d1,0 (ssd1):
   SCSI transport failed: reason 'timeout': giving up
   scsi: [ID 243001 kern.info] /pci@1f,2000/SUNW,ifp@1 (ifp0):
    LIP reset occured; cause f701
   kern.warning] WARNING: /pci@1f,2000/SUNW,ifp@1/ssd@w2000005013b338d1,0
(ssd1):
     LIP occured; cause f701
     scsi: [ID 243001 kern.info] /pci@1f,2000/SUNW,ifp@1 (ifp0):
     LIP reset occured; cause f701
     LIP occured; cause f7e1

     LIP occured; cause f801
     Loop reconfigure in progress
     Unable to allocate target structure for switch setting 4
     reason: Target does not have a hard address
     Loop reconfigure done
     transport rejected (-2)

If you cd onto the disks :
   Read error on the disks

        % mkdir aaa
        mkdir: Failed to make directory "aaa"; I/O error

On the raid, if I run verify on the disks , it runs successfully ...this
determines that the controller
can still talk to the disks but the hostem cannot .

Has anybody experience this problem before?
I have all the latest fibre patches on the host system.

   
   System Configuration on the 420R:
   sun4u Sun Enterprise 420R (2 X UltraSPARC-II 450MHz)
   System clock frequency: 113 MHz
   Memory size: 2304 Megabytes

   Host Bus Adapter on the 420R:
   SUN StorEdge PCI FC-100 Host Adapter ( QLogic QLA2100F)

   On the Raid,
   Open Storage Solutions OmegaFX2 (Chaparral K7313 K410 (e037),
   baselevel K410R03)
   with disk capacity 51GB.

The Raid is connected via a MIA adapter (MIA-1000 1.25 Gb/s DB9 to SC)
 Copper to Optical
 For externally converting a copper I/O to Fiber Optic.

-- 
Seela Balkissoon
Computer Operations Manager
Department of Computer Science
York University
phone: 416 736 2100 x 44324
_______________________________________________
sunmanagers mailing list
sunmanagers@sunmanagers.org
http://www.sunmanagers.org/mailman/listinfo/sunmanagers


This archive was generated by hypermail 2.1.7 : Wed Apr 09 2008 - 23:25:09 EDT