There are various prior questions including.. http://search-hadoop.com/m/4TaT4ts2oz1/disaster+recovery/v=threaded
Is there a clear document on disaster recovery patterns for K and their respective trade offs. How are actual prod deployments dealing with this. For instance I want my topics replicated to another site/cluster and I want the offsets in the topic to match so that a consuming client can fail over. Doors anyone have something like this working really reliably in prod? I can tolerate the reliant being second behind the primary but not differing offsets. What are others managing to achieve with K?
