IBM Storwize V7000 Maintenance Manual page 154

801 • 835
either because of a service assistant user action or
because the node was deleted from the cluster.
User response: When it is no longer necessary to hold
the node in the service state, exit the service state to
allow the node to run:
1. Use the service assistant action to release the service
state.
Possible Cause-FRUs or other:
v None
801 Memory reduced.
Explanation: Memory is reduced but sufficient
memory exists to run I/O operations.
User response: Follow troubleshooting procedures to
fix the hardware.
803 One or more Fibre Channel ports are
not operational.
Explanation: One or more Fibre Channel ports are not
operational.
User response: Follow troubleshooting procedures to
fix the hardware.
1. If possible, use the troubleshooting fix procedures in
the management GUI to correct the associated
cluster error.
2. If no cluster is operational, check whether all Fibre
Channel cables are fully inserted into the node
canister, and that the SAN switch is powered on
and not showing errors. This node error will not
prevent the node from becoming active in a cluster.
If the specified checks do not resolve the problem,
continue with cluster creation and then use the
management GUI to resolve the problem.
805 One or more configured Ethernet ports
are not operational.
Explanation: One or more configured Ethernet ports
are not operational.
User response: Follow troubleshooting procedures to
fix the hardware.
1. If possible, use the troubleshooting fix procedures in
the management GUI to correct the associated
cluster error. If this error has occurred in the
configuration node, you may not be able to run the
management GUI.
2. If you are not able to run the management GUI,
check whether all Ethernet cables are fully inserted
into the node canister, and that the Ethernet switch
is powered on and not showing errors. This node
error will not prevent the node from becoming
active in a cluster. If the specified checks do not
resolve the problem, and you have not created the
cluster yet, continue with cluster creation on the
138
Storwize V7000: Troubleshooting, Recovery, and Maintenance Guide
other node canister in the enclosure and then use
the management GUI to resolve the problem.
3. Use a USB key to get the status of the node and
check whether it is the configuration node. If the
node showing this error is the configuration node,
use a USB key or the service assistant to get the
status of the other node in the enclosure. If the
other node is active, run the service assistant on the
other node, but make this node the active node.
Check whether the active node is reporting node
error 805. Use the service assistant to hold the node
in service state, which will cause the other node in
the enclosure to become the configuration node.
Once you have set the node canister into service
state, you should immediately release the service
state so that the node becomes active in the cluster.
Now use the troubleshooting fix procedures in the
management GUI to correct the associated cluster
error.
815 Cannot determine the VPD for a
component.
Explanation: An FRU in the system has been changed,
and the VPD is unreadable or unrecognized.
User response:
1. Check whether the replacement part that you have
installed is the correct part.
2. See whether there is an updated software package
that correctly supports the part that was used. If an
updated software package exists, upgrade to that
software version. Otherwise, obtain the correct
replacement part for the enclosure model and
software version that you are operating.
820 The node canister has detected that it
has a hardware type that is not
compatible with the control enclosure
MTM, such as node canister type 300 in
an enclosure with MTM 2076-112.
Explanation: This is an expected condition when a
control enclosure is being upgraded to a different type
of node canister. Check that the upgrade instructions
have been followed completely.
User response: If the upgrade instructions have been
followed, this non-critical node error should be
serviced by using the management GUI and running
the recommended actions for the alert with error code
3020.
835 Unable to communicate with partner
node over PCIe link.
Explanation: The partner node is not in a state that
allows communication because the node is powered off,
a boot failure has occurred, or the PCIe link is broken.