Bulk Operation

FLINK-28374 Sorting all unfinished readers in batches at one time in SortMergeResultPartitionReadScheduler

FLINK-28374 Avoid notifying too frequently when recycling buffers for BatchShuffleReadBufferPool

FLINK-28374 Enlarge the max requested buffers for SortMergeResultPartitionReadScheduler

FLINK-28374 Allow multiple downstream consumer job vertices sharing the same intermediate dataset at scheduler side

FLINK-28374 Extract header fields of Buffer into a BufferHeader class for blocking shuffle file IO

FLINK-28374 Store the number of bytes instead of the number of buffers in index entry for sort-shuffle

FLINK-28374 Remove the unused field in SortMergeSubpartitionReader

FLINK-28374 Fix the bug that SortMergeResultPartitionReadScheduler may not read data sequentially

FLINK-28374 Remove data flush in SortMergeResultPartition

FLINK-28374 Select HashBasedDataBuffer and SortBasedDataBuffer dynamically based on the number of network buffers can be allocated for SortMergeResultPartition

FLINK-28374 Introduce new compression algorithms of higher compression ratio

FLINK-28374 Produce one intermediate dataset for multiple consumers consuming the same data

FLINK-28374 Decrease the memory size per request for sort-shuffle data read from 8M to 4M

FLINK-28374 Restrict the number of threads for sort-shuffle data read

FLINK-28374 Read a full buffer of data per file IO read request for sort-shuffle