Bulk Operation

HPCC-27887 Avoid storing compressed data in memory

HPCC-27887 Publish accurate in-memory index sizes

HPCC-27887 Compress blobs using lz4

HPCC-27887 Optimize lookup of a range option

HPCC-27887 Optimize the inplace keybuild code

HPCC-27887 Rename variable to make code clearer

HPCC-27887 Investigate dynamic payload lengths

HPCC-27887 Optimize TLK and fixed length file positions

HPCC-27887 Create test cases for inplace indexes

HPCC-27887 Avoid storing duplicate trailing keyed components

HPCC-27887 Optimize option with range of values

HPCC-27887 Support other non-incremental compression

HPCC-27887 Improve the block-compressed serialization

HPCC-27887 Reuse temporary buffers when building an index

HPCC-27887 Expand inplace payload on demand

HPCC-27887 Reuse the compressor when building and index

HPCC-27887 Document the new compression format options

HPCC-27887 Check that increment in payload size cannot cause internal consistency problems

HPCC-27887 Support non-incremental compression schemes

HPCC-27887 Minor refactoring of inplace index code

HPCC-27887 Fix inplace compression issues revealed by defaulting on

HPCC-27887 Extend dumpkey to support recoding/analysing existing indexes

HPCC-27887 Store the inline payload last

HPCC-27887 Support uncompressed payloads

HPCC-27887 inplace indexes - minor code optimizations

HPCC-27887 Inplace compression improvements

HPCC-27887 Add POC compression format to act as template for new index layouts

HPCC-27887 Refactor jhtree class structure to support additional node types

HPCC-27887 Refactor to support nodes types that can search without decompressing

HPCC-27887 Refactor to allow nodes to support additional compression methods

HPCC-27887 Publish extended index information from Thor

HPCC-27887 Add original blob size details to file meta

HPCC-27887 refactor in preparation for new index formats

HPCC-27887 Display details of index on file information page

HPCC-27887 Store the maximum row length in the header when building an index

HPCC-27887 Default FILEPOSITION(FALSE) on indexes

HPCC-27887 Publish extended information for indexes in eclwatch

HPCC-27887 Extend index read stats to include disk time and #from page cache

HPCC-27887 Integrate index cache with roxiemem

HPCC-27887 Investigate leaf compressing every N items and decompressing on demand

HPCC-27887 add a function to calculate index sizes for a query/queryset

HPCC-27887 Add number of leaf, blob and branch nodes to published meta

HPCC-27887 Add stats for number of leaf/branch/blob nodes in an index write

HPCC-27887 Introduce a inplace compression for leaf nodes

HPCC-27887 Create a new index format for non-leaf nodes.