HP 9000 Virtual Library System User Manual
Page 214
3.
Install the operating system on the new hard drives and restore the VLS. See
.
Recovering from a Primary Node Failure using a Cold Spare Primary Node
On a multi-node VLS, the primary node maintains the configuration for the entire VLS library. In
the unlikely event of a primary node failure, the VLS library would be unavailable until the node
is replaced or repaired. To avoid extended downtime, you can configure an additional node as
a “cold failover” node with all the same attributes as the primary. If the primary node fails, deploy
the spare primary node in place of the failed primary node to bring the entire VLS library back
online quickly.
Install the spare primary node into the rack directly above the primary node.
To deploy the spare primary node:
1.
Shut down all nodes on the VLS.
Because the original primary node has failed, he secondary nodes should be shut down from
the command line of each node. Log in to the administrator account on each secondary node
using the serial port or keyboard and mouse. At the prompt, enter shutdownNode.
2.
Cable the spare primary node exactly as the original primary node was cabled:
a.
Move all Fibre Channel cables from the original primary node ports to the corresponding
spare primary node ports.
b.
Move all Ethernet cables including the USB Ethernet adapter from the original primary
node ports to the corresponding spare primary node ports.
c.
Connect to the serial port or use the keyboard and mouse ports to connect to the console
of the spare primary node.
3.
Power on the spare primary node.
The node will run cable checks and configuration checks. After several minutes, a menu will
appear on your monitor asking whether the node is a primary (master, m) or secondary (slave,
s) node.
214
Disaster Recovery