IBM Storwize V7000 Maintenance Manual Download

Page: 58 / 172

The failure is reported regardless of the method that you used to create a clustered
storage system:

USB key

Service assistant

Service command line

The create clustered-system function protects the system from loss of volume data.
If you create a clustered system on a control enclosure that was previously used,
you lose all of the volumes that you previously had. To determine if there is an
existing system, use data that is returned by “Procedure: Getting node canister and
system information using the service assistant” on page 48 or “Procedure: Getting
node canister and system information using a USB key” on page 49.

The node canister that you are attempting to create a clustered system on is in
candidate state. The node canister is in candidate state if it is a new canister.

The partner node canister in the control enclosure is not in active state.

The latest system ID of the control enclosure is 0.

If the create function failed because there is an existing system, fix the existing
clustered system; do not re-create a new clustered system.If you want to create a
clustered system and do not want to use any data from the volumes used in the
previous clustered system, go to “Procedure: Deleting a system completely” on
page 56, and then run the create function again.

You might not be able to create a cluster if the node canister (the one on which
you are attempting to create the clustered system) is in service state. Check
whether the node canister is in service state by using the data returned by
“Procedure: Getting node canister and system information using the service
assistant” on page 48 or “Procedure: Getting node canister and system information
using a USB key” on page 49. If the node is in service state, fix the reported node
errors. For more information, go to “Procedure: Fixing node errors” on page 56.
After the node error is corrected, attempt to create a clustered storage system
again.

Problem: Unknown service address of a node canister

This topic describes the methods that you can use to determine the service address
of a node canister.

If you can access the management GUI, start the service assistant on the
configuration node for the clustered system by pointing your web browser to the
address:

control enclosure management IP address

/service. For example, if your

control enclosure management IP address is 11.22.33.44, point your web browser
to 11.22.33.44/service.
The service assistant home page lists the node canister within the clustered
system that it can communicate with. If the service address of the node canister
that you are looking for is listed in the

Change Node

window, make the node

the current node. Its service address is listed under the

Access

tab of the node

details. Use the service address to connect to the node or continue to manage the
node using this session. If the service address of the node that you want is not
listed, go to “Procedure: Getting node canister and system information using a
USB key” on page 49 to get the service address.

If you know the service address of any node canister in the system, follow a
similar procedure to the one described previously. Rather than using

control

enclosure management IP address

/service to start the service assistant, use the

service address that you know.

Storwize V7000: Troubleshooting, Recovery, and Maintenance Guide

|
|
|
|
|
|

|
|
|
|
|

|
|
|
|
|
|
|
|
|

|
|
|
|
|

|
|
|
|
|
|
|
|

|
|
|
|