You can configure an Adaptive Server resource group to withstand multiple node failures by configuring multiple secondary nodes. All the nodes that can potentially host the Adaptive Server resource group are specified in the resource group property Nodelist.
For example, to create a resource group with multiple nodes, use:
scrgadm -a -g rg_MONEY -y Nodelist=node1,node2,node3
The order of the node names is the preference in which the Resource Group Manager selects a node to host the Adaptive Server resource group. When Adaptive Server fails over, it does so to the next available secondary node, as determined by the Sun Cluster Resource Group Manager.
As long as at least one of the potential primary nodes is available, the Adaptive Server resource group is available, regardless of the number of crashes.
Figure 10-3 describes a three-node setup.
In this example, the Adaptive Server running on Node1 could fail over to either Node2 or Node3, depending on the order of the node list. Because Node2 crashed, Adaptive Server fails over to Node3. If other nodes are included in the node list as candidates for secondary nodes, Adaptive Server can fail over to any of them as well.
After the primary node is brought back online, you can either fail back to it, or keep it available as a candidate for a secondary node.
| Copyright © 2005. Sybase Inc. All rights reserved. |
|
|
