10.0 Deploying eDirectory on High Availability Clusters

The primary method through which NetIQ eDirectory supports high availability is by configuring multiple servers through synchronization. However, clustering may be a more viable alternative for achieving high availability in some environments.

This section provides guidelines for configuring eDirectory on high availability clusters by using shared storage. The information in this section is generalized for shared storage high availability clusters on supported Windows and Linux platforms, and the information is not specific to a particular cluster manager.

State data for eDirectory must be located on the shared storage so that it is available to the cluster node that is currently running the services. This means that the eDirectory DIB must be located on the cluster shared storage. The root eDirectory instance on each of the cluster nodes must be configured to use the DIB on the shared storage.

In addition to the DIB, it is also necessary to share NICI (NetIQ International Cryptographic Infrastructure) data so that server-specific keys are replicated among the cluster nodes. NICI data used by all cluster nodes must be located on the cluster shared storage.

Other eDirectory configuration and log data should also reside on shared storage.

eDirectory 9.1 includes a utility for both Linux and Windows servers that automatically configures eDirectory in your clustered environment, including copying data to a specified shared storage location, updating the appropriate configuration parameters, and setting up eDirectory services on cluster nodes other than the primary node.

The procedures in the following sections are based on the following assumptions:

  • You are familiar with eDirectory installation procedures.

  • You are using a two-node cluster.

    NOTE:A two-node cluster is the minimum configuration used for high availability. However, the concepts in this section can easily be extended to a cluster with additional nodes. Note that eDirectory does not support load balancing by using multiple cluster nodes.

This section covers the following topics: