Class ShortReadProtection
- java.lang.Object
-
- org.apache.cassandra.service.reads.ShortReadProtection
-
public class ShortReadProtection extends java.lang.Object
We have a potential short read if the result from a given node contains the requested number of rows (i.e. it has stopped returning results due to the limit), but some of them haven't made it into the final post-reconciliation result due to other nodes' row, range, and/or partition tombstones. If that is the case, then that node may have more rows that we should fetch, as otherwise we could ultimately return fewer rows than required. Also, those additional rows may contain tombstones which which we also need to fetch as they may shadow rows or partitions from other replicas' results, which we would otherwise return incorrectly.
-
-
Constructor Summary
Constructors Constructor Description ShortReadProtection()
-
Method Summary
All Methods Static Methods Concrete Methods Modifier and Type Method Description static UnfilteredPartitionIterator
extend(Replica source, java.lang.Runnable preFetchCallback, UnfilteredPartitionIterator partitions, ReadCommand command, DataLimits.Counter mergedResultCounter, long queryStartNanoTime, boolean enforceStrictLiveness)
-
-
-
Method Detail
-
extend
public static UnfilteredPartitionIterator extend(Replica source, java.lang.Runnable preFetchCallback, UnfilteredPartitionIterator partitions, ReadCommand command, DataLimits.Counter mergedResultCounter, long queryStartNanoTime, boolean enforceStrictLiveness)
-
-