[
https://issues.apache.org/jira/browse/IGNITE-26018?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=18009592#comment-18009592
]
Ivan Bessonov commented on IGNITE-26018:
----------------------------------------
Before:
{code:java}
Benchmark (columnTypes) Mode Cnt Score
Error Units
SortedIndexTreeInsertBenchmark.putx LONG thrpt 10 478.734 ±
11.138 ops/ms
SortedIndexTreeInsertBenchmark.putx STRING_16 thrpt 10 337.458 ±
46.529 ops/ms {code}
After:
{code:java}
Benchmark (columnTypes) Mode Cnt Score
Error Units
SortedIndexTreeInsertBenchmark.putx LONG thrpt 10 669.271 ±
76.296 ops/ms
SortedIndexTreeInsertBenchmark.putx STRING_16 thrpt 10 486.232 ±
22.141 ops/ms {code}
> POC for runtime-generated binary tuple comparator
> -------------------------------------------------
>
> Key: IGNITE-26018
> URL: https://issues.apache.org/jira/browse/IGNITE-26018
> Project: Ignite
> Issue Type: Improvement
> Reporter: Ivan Bessonov
> Assignee: Ivan Bessonov
> Priority: Major
> Labels: ignite-3
> Time Spent: 10m
> Remaining Estimate: 0h
>
> Binary tuple comparators suffer from too many virtual calls and too many
> branching. Given that index schema is immutable, we can use this information
> for our advantage and compile special optimized comparator for each index
> schema.
> After the POC is complete we can use
> {{org.apache.ignite.internal.storage.pagememory.benchmarks.SortedIndexTreeInsertBenchmark}}
> for testing the difference.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
