During the past week we've failover several of our mongo clusters. After the failover happened, we were unable to perform any writes - each one failing with "Mongo::Error::OperationFailure (not master and slaveOk=false)". This issue was solved only after redeploying our application.

We've specified all of the hosts in our mongoid.yml file but still Mongoid doesn't seem to be able to connect to the new master.

Any ideas on what might be causing this / suggestions for how to fix it?