Server Group Awareness (SGA), explanation enhancements

The current documentation page mentions the following in SGA general notes:

"Server Group Awareness only applies to the Data Service."

Furthermore, the article doesn't explain SGA distribution of any other services (index, query, etc.). This can be misleading for the customers. We need to explain that for group failover to work properly on a production cluster, all services on that cluster need to be planned accordingly. E.g., if there are three server groups in the production cluster, each group should have index, query, and other services running on the nodes in the corresponding racks/AZs. If all index nodes run in the same AZ, and that AZ fails, their presence in different server groups will not help. This may be obvious to us, but it's not so much to the customers. Along these lines, it would help to have a diagram that shows not only data nodes distributed over server groups but the nodes with other services, too.

As a minor change, shouldn't we say "in the group" instead of "on the group" everywhere in the text?

Reporter: Oleg Kuzmin
E-mail: oleg.kuzmin@couchbase.com