A contrast between mainframe Parallel Sysplex availability and selected distributed computing availability solutions

Critical applications worldwide are deployed on mainframes, and a chief reason for this is the mainframe infrastructure for high availability. In this paper, we describe the history and motivation behind GDPS®, the IBM Geographically Dispersed Parallel Sysplex, and we contrast GDPS with availability solutions that have been deployed on distributed computing platforms at Yahoo!, Amazon, and Google.