What replication scenarios are supported by RavenHQ?

This is the default replication scenario that is setup for all replicated plans on RavenHQ. Master/slave will give you failover for all read operations without any special configuration in the RavenDB client. If the master node goes down, write operations will not be able to succeed.

 

We can configure master/master replication for your database if you send a support ticket to support@ravenhq.com.  Master/master allows your database to failover on both reads and writes, but it requires you to change your RavenDB client's FailoverBehavior from the default (AllowReadsFromSecondaries) to AllowReadsFromSecondariesAndWritesToSecondaries. For more information on this subject, please review the RavenDB Client failover documentation. 

 

Because master/master replication introduces the possibility of conflicts, we need to consider a strategy for resolving those conflicts. In RavenHQ we automatically setup a "last-wins" conflict resolver for all version 2.5 databases. This means that in the event of a conflict, the most recently updated document will be persisted and it will overwrite any information in the older document. In RavenDB 3.0, you are able to select your own conflict resolution strategy directly in the RavenDB Management Studio. 

 

If you have any additional questions, drop us a line at support@ravenhq.com.