Class BatchedDefinitionResolver<N extends HistoricalNode & Comparable<? super N>>

java.lang.Object
com.inductiveautomation.historian.gateway.api.storage.AbstractDefinitionResolver<N>
com.inductiveautomation.historian.gateway.api.storage.BatchedDefinitionResolver<N>
Type Parameters:
N - the node type the backend stores; see AbstractDefinitionResolver.

public abstract class BatchedDefinitionResolver<N extends HistoricalNode & Comparable<? super N>> extends AbstractDefinitionResolver<N>
Batching layer over AbstractDefinitionResolver. Definition writes are buffered per flush and written in one round trip, then published to the lookup caches only after they are durable.

Callers validate points with validatePointForBatch(P, BatchContext, BatchedDefinitionResolver.DefinitionBatch), then call flush(BatchedDefinitionResolver.DefinitionBatch, BatchContext) (or run executeDefinitionBatch(List) themselves and call stageForCachePublication(List, BatchedDefinitionResolver.FlushResult, BatchContext, BatchedDefinitionResolver.DefinitionBatch)), and finally drainPendingCachePublications(BatchedDefinitionResolver.DefinitionBatch, BatchContext) after the value writes commit, or invalidateStagedKeys(BatchContext) on rollback. Backends implement executeDefinitionBatch(List), withResolvedId(N, HistoricalNodeId), publishResolvedNode(N), pathKeyFor(N) and isImplicit(N).

Safe for concurrent validation threads: per-point scratch state is thread-local and the batch they buffer into is deduplicated by path.