I questioned the decision of a colleague concerning the use of a HOSTS file entry on an application server (standard SQL database backed .NET solution), which I think would cause connections to the SQL Cluster to connect only a single node in the SQL Cluster. In other words, the by using a HOSTS file entry like this....
10.10.0.100 SQLCLUSTER
...when the IP address above is the IP address of the active node in the SQL cluster and the application connection string contains SQLCLUSTER as the data source.
Intuitively, I would think the clustering technology would fail in an Active-Passive cluster configuration when using HOSTS files as described above. I do not have in-depth knowledge of MS SQL clustering, just more traditional network load balancing. Am I correct?