Procedure: Deleting A Cluster Completely; Procedure: Fixing Node Errors - IBM Storwize V7000 Troubleshooting And Maintenance Manual

The status, speed, and MAC address are returned for each port. Information is returned that identifies
whether the node is the configuration node and if any node errors were reported.
v Examine the two LEDs of the Ethernet ports.
The activity LED flashes when there is activity on the connection. The link state LED must be
permanently on. If it is off, the link is not connected.
Until the Ethernet link is connected, perform the following actions to check the port status each time:
1. Verify that each end of the cable is securely connected.
2. Verify that the port on the Ethernet switch or hub is configured correctly.
3. Connect the cable to a different port on your Ethernet network.
4. If the status is obtained using the USB key, review all the node errors that are reported.
5. Replace the Ethernet cable.
6. Follow the hardware replacement procedure for a node canister. See "Replacing a node canister" on
page 85.

Procedure: Deleting a cluster completely

This procedure guides you through the process to completely remove all cluster information from a
system. When the procedure is finished, the system performs like a new installation.
CAUTION:
This procedure makes all the volume data that you have on your system inaccessible. You cannot
recover the data. This procedure affects all volumes that are managed by your system, which includes
the drives in the control enclosure, in the expansion enclosures, and on the managed disks on external
storage systems. The only volumes not affected are image mode volumes on external storage systems.
Do not continue unless you are certain that you want to remove all the volume data and configuration
data from your system. This procedure is not used as part of any recovery action.
There are two stages to this procedure. First, the node canisters are reset. Second, the enclosure data is
reset.
1. Start the service assistant on one of the node canisters.
2. Use the service assistant node action to hold the node in service state.
3. Use the Manage Cluster option to remove the cluster data from the node.
4. Perform the previous three steps on the second node canister in the enclosure.
Note: Instead of connecting to the other node, you can make the second node the current node.
5. Open the service assistant Configure Enclosure and select Reset Cluster ID option.
This action causes the system to reset.
6. Power the system off and on before creating a cluster.

Procedure: Fixing node errors

This procedure describes how to fix a node error that is detected on one of the node canisters in your
system.
Node errors are reported when there is an error that is detected that affects a specific node canister.
1. Use the service assistant to view the current node errors on any node.
2. If available, use the management GUI to run the recommended action for the alert.
3. Follow the fix procedure instructions.
Chapter 5. Resolving a problem
45