Package org.apache.spark.storage
package org.apache.spark.storage
-
ClassDescriptionAbstracts away how blocks are stored and provides different ways to read the underlying block data.:: DeveloperApi :: Identifies a particular Block of data, usually associated with a single file.:: DeveloperApi :: This class represent a unique identifier for a BlockManager.The response message of
GetLocationsAndStatusrequest.Driver to Executor message to get a heap histogram.Driver to Executor message to trigger a thread dump.::DeveloperApi:: BlockReplicationPrioritization provides logic for prioritizing a sequence of peers for replicating blocks.:: DeveloperApi :: Stores information about a block status in a block manager.Helper class that ensures a ManagedBuffer is released upon InputStream.close() and also detects stream corruption if streamCompressedOrEncrypted is trueA TopologyMapper that assumes all nodes are in the same rackA simple file based topology mapper.Helper class forShuffleBlockFetcherIteratorthat encapsulates all the push-based functionality to fetch push-merged block meta and shuffle chunks.A listener to be called at the completion of the ShuffleBlockFetcherIterator param: data the ShuffleBlockFetcherIterator to process:: DeveloperApi :: Flags for controlling the storage of an RDD.A mapper class easy to obtain storage levels based on their names.Helper methods for storage-related objects.Intercepts write calls and tracks total time spent writing in order to update shuffle write metrics.::DeveloperApi:: TopologyMapper provides topology information for a given host param: conf SparkConf to get required properties, if needed