LoadBalancingConfig
Load balancing policy for multi-datacenter deployments. dcAwareFailover, allowedRemoteDcs, and FailoverConfig.onLocalDcUnavailable are wired into the DataStax driver's own DefaultLoadBalancingPolicy (GH #58) -- setting dcAwareFailover = true together with failover { onLocalDcUnavailable = FailoverPolicy.RETRY_REMOTE_DC } genuinely enables the driver's cross-DC failover mechanism, restricted to the datacenters listed in allowedRemoteDcs. Setting only one of the two leaves failover inert, by design (see FailoverPolicy).
Properties
DCs eligible for failover. Required (non-empty) when dcAwareFailover = true. Enforced as an allow-list via a driver NodeDistanceEvaluator (GH #58) -- every other remote DC is kept at NodeDistance.IGNORED and never connected to. The driver does not expose a priority/ordering mechanism across multiple allowed remote DCs; all of them are equally eligible once enabled.
Allow the driver to use replicas in remote DCs when the local DC is unavailable. Must be combined with failover { onLocalDcUnavailable = FailoverPolicy.RETRY_REMOTE_DC } to actually take effect -- see FailoverPolicy.
Maximum number of remote replicas used per remote DC during failover. Wired directly to the driver's advanced.load-balancing-policy.dc-failover.max-nodes-per-remote-dc option (GH #58) whenever failover is actually enabled (see FailoverPolicy).
Route queries to the token owner — avoids coordinator hop (always recommended). The driver's DefaultLoadBalancingPolicy is token-aware unconditionally by design (there is no driver-level toggle to disable it without switching to a materially different policy), so this flag currently only documents the recommended posture rather than being read anywhere.