IBM Storwize V7000, Problem Determination Manual

Page: 56 / 314

In isolating problems, be sure to review the labels on the rear of the systems for
exact port plugging.
Software detected problems via event codes:
If you have been directed to this procedure by a software event code, use the
Monitoring > System page in the management GUI to identify the effected file
module or refer to the following procedure to determine the logical to physical
mapping of the event then proceed to the physical hardware isolation procedures.
The isolation of Fibre Channel connections based on a single error event is not
simple. As Figure 3 on page 31 shows, there are two file modules attached to the
control enclosure; however, the logical host name of these systems does not map
directly to the connections. The logical host name of the file module depends on
which file module is used for initial USB key installation. For example, in Figure 3
on page 31 file module
B
can have a host name of
mgmt001st001
if the
installation was initiated on that node or it might have a host name of
mgmt002st001
if the installation was initiated on the second file module. Each error
event is reported against the logical host name where the problem occurred.
For isolation of Fibre Channel connections, it is important with a single file module
that both Fibre Channel connections go to the same port number on both the
upper and lower Storwize V7000 node canisters. Port 1 always goes to the upper
canister, and port 2 goes to the lower canister.
Use the table for correlating the error code with the physical connections and
follow the procedures after the table for enabling the LED indicator on the front of
the file module.
Table 11. Error code port location mapping
Error code Description
File Module Fibre
Channel Location
Storage Node
Canister Fibre
Channel Port
4B0800C Link failure. Fibre
Channel adapter 1,
port 1 not up. PCI slot #2 – port 1
(right port when
facing rear of system) Upper node canister,
port 1. OR Upper
node canister, port 2.
4B0801C Link failure. Fibre
Channel adapter 1,
port 2 not up. PCI slot #2 – port 2
(left port when facing
rear of system) Lower node canister,
port 1. OR lower
node canister, port 2.
4B0803C Slow connection on
Fibre Channel
adapter 1, port 1. PCI slot #2 – port 1
(right port when
facing rear of system) Upper node canister,
port 1. OR upper
node canister, port 2.
4B0804C Slow connection on
Fibre Channel
adapter 1, port 2. PCI slot #2 – port 2
(left port when facing
rear of system) Lower node canister,
port 1. OR lower
node canister, port 2.
To enable the LED indicator for the node reporting the problem use the
Monitoring > System page in the management GUI or follow this procedure:
1. Log onto the active file module via the CLI interface.
2. Run the command:
cnlocatenode #HOSTNAME on #SECONDS
.
HOSTNAME
is the
hostname associated with the error... either
mgmt001st001
or
mgmt002st001
.
#SECONDS
is the number of seconds for the LED indicator to be turned on.
Physical connection and repair:
32
Storwize V7000 Unified: Problem Determination Guide Version