High availability with shared disk

Many highly available configurations require a shared file system to host common installation locations, data, and instances. You can also implement a shared file system to store attachments, indexes, and error logs that are used in an integration framework scenario.

You can configure shared disks and shared disk clusters to be highly available through redundancy. Redundant shared disks can integrate easily with your high availability environment.

When a shared disk is used, it is important that it does not represent not a single point of failure. Various methods exist to achieve highly available file systems. For example, you can use IP-based replication, local mirroring, clustered network-attached storage (NAS), or a clustered storage area network (SAN).

You can set up shared disks by configuring them to be simultaneously mounted on multiple servers in an active-active configuration or to be auto-mounted on the active node in a warm-standby configuration.

Choose from one of the following methods to configure a shared disk:

Auto-mounting

Auto-mounting is a method that configures shared disks by using a specialized software program. With auto-mounting, disks are mounted only as they are accessed and are removed after a period of inactivity. To ensure high availability, you must ensure that auto-mounted file systems do not time out.

Manual mounting

Manual mounting provides read and write access to files from multiple servers. For detailed information about manual mounting, see the documentation for your operating system and the documentation for the shared disk that you use.

Mounting with System Automation for Multiplatforms

You can use System Automation for Multiplatforms to configure and control the mounting of shared disks in the environment. To ensure high availability for all of the shared data, the directories that contain the data are shared on a highly available disk and mounted on a primary node. During a system failure, System Automation for Multiplatforms removes the directories from the primary node and mounts them on a standby node. Before you use System Automation for Multiplatforms for your shared disk implementation, check that this solution is fully compatible with your operating system.