In some instances, the switch port may end up stuck in this state, depending
on the failure mechanism of the device on that port, such as the device
rebooting repeatedly due to power problems, and so on, and the port can
be individually reset by entering the
portDisable
command followed by
the
portEnable
command. In rare instances, you can remotely reboot the
entire switch by using the
reboot
command, which takes approximately one
minute to complete. This will cause all of the Fibre Channel devices (systems
and HSG80 ports) to log out and log back in, but no system or HSG80 reboots
will be required. Depending on the ASE configuration parameters, services
will be relocated or go off line, and will require relocation or restarting to
get them back on line and operational. These mechanisms are also available
on the GUI interface.
You can monitor cluster I/O performance using the
portPerfShow
command. This command displays the throughput (in bytes/sec, KB/sec, or
MB/sec) on each switch port in use and updates its output each second. The
GUI interface displays a graph of the throughput on each port in real time.
For more information on managing the switch, see the SAN Switch Fabric
Operating System Management Guide.
5.1.3 Fibre Channel Hub Failures
The Fibre Channel hub represents a single point of failure (SPOF) for the
SCSI bus that it serves. A total failure of one hub will cause the entire
shared bus to fail. Depending on how the storage is configured, this failure
can result in the loss of access to some or all of the data. If the cluster has
two hubs, as shown in Figure 1–8, and the critical data is mirrored using
the Logical Storage Manager (LSM) across the two buses, then there would
be no interruption and you could resynchronize the data mirror after the
hub is serviced.
5.1.4 HSG80 RAID Array Controller Failures
Each HSG80 storage array (RA8000 or ESA12000) consists of one or two
controllers. In a dual-redundant configuration with transparent failover
mode enabled, one left port and one right port are active at any given time.
If the top controller fails in such a way that it can no longer properly
communicate with the switch, then its functions will automatically fail over
to the bottom controller (and vice versa).
Because the HSG80 controller’s configuration information and worldwide
name is stored in nonvolatile random-access memory (NVRAM) on the
controller, there are different procedures for replacing HSG80 controllers
in an RA8000 or ESA12000:
Troubleshooting and Component Replacement Guidelines 5–3
