Reputation: 19
We have implemented a four node Cassandra cluster in AWS and now wish to migrate the same data to Azure.currently when we are trying to do Cassandra replication we observed it is in a active active mode and load get splits among all servers of the primary and secondary sites this is leading to data inconsistency in results for queries made on the primary site as it somehow balances the load and gets data rendered from the secondary site we would like your help to know if that can be restricted or there is a fail over sync possible in Cassandra. Any suggestions shall be much appreciated. Thanks
Upvotes: 0
Views: 149
Reputation: 339
Try DCAwareRoundRobinPolicy and specify .withLocalDc("Your Data Center to connect") and use Consistency of Local_one or Local_quorum.
Upvotes: 0
Reputation: 3938
You want to use the NetworkTopologyStrategy
for replication then your app code should do LOCAL_ONE
consistency etc. As for the failover you can have your apps reboot configured to point at a particular DC. Or you could have the secondary site have its own servers ready that you fail over to the whole site.
Upvotes: 2