11.7 Summary

High availability begins with defining the failure types for the application solution for which recovery is required, the speed with which the recovery should happen, and the various hardware, middleware, and application parts that constitute the high-availability solution.

The mainframe already has multiple levels of built-in hardware redundancy that contribute to its intrinsic high availability. You can configure the hardware so that the software can directly exploit various levels of hardware redundancy, such as the CPUs and channel paths to devices. Using z/VM virtualization, you can create various types of Linux configurations that provide software redundancy to an application. For example, you can configure hot-standby Linux images without investing in additional hardware.

It is possible today to implement some high-availability solutions with the existing Linux-on-the-mainframe tools. Besides developing their own high-availability products, IBM and other vendors are also supporting several Open Source projects that are working to further improve the high-availability tool set. Additional full-blown products are likely to become available during the year 2003.