[ https://issues.apache.org/jira/browse/IGNITE-14743?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17370514#comment-17370514 ]
Andrey Mashenkov commented on IGNITE-14743: ------------------------------------------- PR with different approaches: * Current approach with minor optimizations [1]. * Precalculate row size then marshall tuple with the preselected format [2]. * Marshall row, then compact if needed [3]. PFA benchmark results. [1] https://github.com/apache/ignite-3/pull/171 [2] https://github.com/apache/ignite-3/pull/168 [3] https://github.com/apache/ignite-3/pull/188 > Support Row with large values. > ------------------------------ > > Key: IGNITE-14743 > URL: https://issues.apache.org/jira/browse/IGNITE-14743 > Project: Ignite > Issue Type: Improvement > Reporter: Andrey Mashenkov > Assignee: Andrey Mashenkov > Priority: Major > Labels: iep-54, ignite-3 > Fix For: 3.0.0-alpha3 > > Attachments: Byte array columns only benchmark.txt, Fixlen cols only > benchmark.txt, Latin1 string columns benchmark.txt, Non-latin string columns > benchmark.txt, String marshalling comparison > > Original Estimate: 168h > Time Spent: 1h 20m > Remaining Estimate: 166h 40m > > For now, TupleAssembler writes offsets for varlen columns as 2-byte \{{short > }}type. > This implicitly restricts key/value sizes down to 64 kB in total. > Let's use 4 byte types (byte, short, int) for offsets. -- This message was sent by Atlassian Jira (v8.3.4#803005)