I guess you are not right.. Cluster.builder.addContactPoint(...) will add nodes in the connection pool.. And it will discover all the other nodes in the connection pool automatically.. To filter out nodes only for local colo we need to use to different settings in Java driver..
There should be similar stuff in cpp driver as well.. On Tue, Mar 4, 2014 at 8:20 AM, Manoj Khangaonkar <khangaon...@gmail.com>wrote: > Hi , > > Your client/application will connect to one of the nodes from the nodes > you tell it to connect. In the java driver this is done by calling > Cluster.builder.addContactPoint(...). I suppose the C++ driver will have > similar class method. For the app in DC1 provide only nodes in DC1 as > contact points. > > regards > > > On Tue, Mar 4, 2014 at 6:47 AM, Check Peck <comptechge...@gmail.com>wrote: > >> I have couple of question on Datastax C++ driver. >> >> We have 36 nodes Cassandra cluster. 12 nodes in DC1, 12 nodes in DC2, 12 >> nodes in DC3 datacenters. >> >> And our application code is also in three datacenters- 11 node in DC1, 11 >> node in DC2, 11 node in DC3 datacenter. >> >> So my question is if the application call is coming from DC1 datacenter, >> then will it go to DC1 Cassandra nodes automatically with the use of cpp >> driver? And same with DC2 and DC3? >> >> Or we need to add some config changes in our C++ code while making >> connection to cassandra which will then make sure if the call is coming >> from DC1 datacenter then it will go to DC1 Cassandra nodes? >> >> If there is any config change which we need to add in our C++ code, then >> can you please point me to that? >> >> > > > -- > http://khangaonkar.blogspot.com/ >
