Carl Trieloff wrote:

I believe that the simple way is to look at clustering for scalability and for fault tolerance as separate user decisions.

Simplistically
a.) if every broker can be configured to have a passive companion for FT
b.) then a scalability cluster can be made up of single brokers (Scalable not FT) or active-passive pairs ( FT + Scalable)

Thats the approach sonic take also AFAIK.

Reply via email to