|
This document arises because there is an obvious need for it. If nothing else, it's existence will hopefully allow the authors to spend more time doing what they are supposed to (or want to) be doing rather than sorting out performance problems with High Availability Geographic (HAGEO) clusters! What we are trying to achieve by this document is to disseminate some of the knowledge and experience we have gained over the years with HAGEO and GeoRM allowing for the better design and installation of future systems. We see, or are involved with, a lot of actual and potential customer systems. As a result, most of the examples and experiences are derived from work performed on real customer configurations. No explicit customer details are discussed however; customer confidentiality has been preserved throughout. We have also undertaken probably the most intensive set of geomirroring performance tests ever carried out. We will share the results of these tests, derive some design and operational 'rules of thumb', and show the effect of their application on a real customer example configuration.
Throughout this document, we use HAGEO and GeoRM interchangeably except where specifically stated. They utilise the same technology and demonstrate the same performance characteristics.
This document assumes a reasonable knowledge of HAGEO and/or the Geographic Remote Mirror (GeoRM) products. If you don't know what HAGEO or GeoRM are, or what they do, then maybe starting here, is not necessarily the best place! A good introduction with a fair degree of detail may be found in their respective 'Concepts and Facilities' manuals. However, for those of you wishing to push the envelope straightaway, here's a very quick no-holds-barred introduction.
| 
