Sun StorEdge[TM] RAID Manager 6.22.1 software: Device structure doesn't get created, and the "lad" command doesn't show LUNs after a power cycle of A3500 or A1000

Asset ID:	1-72-1002007.1
Update Date:	2009-03-01
Keywords:

Solution Type Problem Resolution Sure

Solution 1002007.1 : Sun StorEdge[TM] RAID Manager 6.22.1 software: Device structure doesn't get created, and the "lad" command doesn't show LUNs after a power cycle of A3500 or A1000

Related Items


Sun Storage A1000 Array
 Sun Storage A3500 SCSI Array
 Sun Storage A3500 FC Array

Related Categories


GCS>Sun Microsystems>Storage - Disk>Modular Disk - Other

PreviouslyPublishedAs
202799

Symptoms
An A1000 or A3500 array with the host running Sun StorEdge[TM] RAID Manager
6.22.1 software is rebooted due to an outage. After the array reboots, the
following has occurred:

* The logical unit numbers (LUNS) do not appear in the output of the "lad"
command.
* The  device structure of the LUNS is not created in the Solaris[TM]
system.
* Output from the "healthck-a" command is a "controller not responsive"
message.

Resolution
When this problem occurs, the following section appears in the "save_module_profile" output of the RAID Manager software for this module. With RAID Manager software, you capture save_module_profile" in the configuration window of the graphical user interface (GUI) by choosing File, and then pressing the "save_module_profile" button:

   Controllers:
Name          Serial Number    Mode           Logical Units
A             1T00311602       Active         6
B (c10t4d1)   1T94601102       Active         5
---
------
Device  RAID  Cap.
LUN   Group  Name    Level (MB) Status
0     1              5   34389  Optimal
1     2    c10t4d1   5   34389  Optimal
2     3              5   34389  Optimal
3     4    c10t4d3   5   34389  Optimal
----
-------
In the preceding example, two LUNs each are assigned to controller A and
B.  Controller B (c10t4d1) appears to be working properly and LUNs 1 and
3 are in Optimal status.  The LUNs assigned to controller A also show as
Optimal, but they don't have any c#t#d# associated with them.
A power cycle of the array and reboot -r on the host can fix this problem,
but you can avoid that process by running the "hot_add" command
to have the system rebuild the device structure.  After a successful
"/usr/lib/osa/bin/hot_add," the output (for the preceding example) should
appear as follows:
Controllers:
Name          Serial Number    Mode           Logical Units
A (c4t0d0)    1T00311602       Active         2
B (c10t4d1)   1T94601102       Active         2
---
------
Device  RAID  Cap.
LUN   Group  Name    Level (MB) Status
0     1    c4t0d0    5   34389  Optimal
1     2    c10t4d1   5   34389  Optimal
2     3    c4t0d0    5   34389  Optimal
3     4    c10t4d3   5   34389  Optimal
----
-------

Product
RAID Manager 6.22.1
Sun StorageTek A3500 FC Array
Sun StorageTek A3500 Array
Sun StorEdge A1000 Array

raid manager, hot_add, lad, unresponsive controller, A3500, A1000, healthck
Previously Published As
75255

Change History
Date: 2006-01-19
User Name: 25440
Action: Update Canceled
Comment: *** Restored Published Content *** SSH AUDIT
Version: 0
Date: 2006-01-19
User Name: 25440
Action: Update Started
Comment: SSH AUDIT
Version: 0
Date: 2004-04-19
User Name: 29589
Action: Approved
Comment: Qlty check - changed the word "are" to "and", removed an extra word "that".
Republish.
Version: 0

Attachments

This solution has no attachment