422 IBM Midrange System Storage Hardware Guide
Use a text editor such as Notepad to view the Capture State Information file. The only
command we need to check is spmShow from each controller. The sections we need to search
for begin with:
Executing spmShow(0,0,0,0,0,0,0,0,0,0) on controller A
and
Executing spmShow(0,0,0,0,0,0,0,0,0,0) on controller B
From there, we need to page down to a subsection with the following heading:
---I-T-Nexus (PORT) Objects (ITN)---
This provides a list of initiator devices that are currently seen by this controller. The HBA ports
can be recognized by either the alias name, which matches the alias defined in the Storage
Manager mappings view or Host_xxxxxxxxxxxxxxxx, where xxxxxxxxxxxxxxxx represents the
WWPN of the undefined HBA. For each Dual-HBA host configured in the recommended way
shown in Figure 7-63 on page 421, we expect to see one entry for each HBA on each
controller. The following output shows an example of how spmShow can be used to
troubleshoot host-side problems:
Controller A spmShow output:
---I-T-Nexus (PORT) Objects (ITN)---
ITNID InitiatorPort TargetPort Online
----- ------------------------------ ------------------------------ ------
x0012 X3655A_port0 FC_TargetPort_Ah_ch11 on
x0012 X3655A_port1 FC_TargetPort_Ah_ch12 off
Controller B spmShow output:
---I-T-Nexus (PORT) Objects (ITN)---
ITNID InitiatorPort TargetPort Online
----- ------------------------------ ------------------------------ ------
x0012 X3655A_port0 FC_TargetPort_Ah_ch11 on
x0012 X3655A_port1 FC_TargetPort_Ah_ch12 off
The Online column shows a status of either on or off:
on Device is currently accessible from this controller.
off Device was recognized on this channel previously but is no longer
accessible.
The output confirms that we need to focus on the HBA named X3655A_port1 for the root
cause of the logical drive moving onto the non-preferred path. This might be due to a zoning
or configuration change, a hardware failure on the HBA card itself, or the fiber link to it.
Checking the host logs and switch port status will reveal more information.
Note: This same technique can be used during initial configuration to confirm that all HBAs
are zoned correctly. It is also good practice to simulate link failures during installation in
order to verify that path failover is configured correctly on each host. Often,
misconfiguration is only detected following a genuine failure when path failover does not
behave as expected.
To Next Page
Loading...
Need help?
General