fallintoplace opened a new pull request, #2226: URL: https://github.com/apache/iceberg-go/pull/2226
## What - Reuse partition results for adjacent rows with equal source values. ## Why - Clustered writes repeat the same transforms and partition lookups. ## Implementation - Cache the last source values and append rows directly while the partition key stays equal. - Keep the current path for nested and unsupported types. ## Benchmark - Go 1.26.3, Apple M1 Pro, 65,536 rows; median of 5 runs. - Runs of 8, 256, and 4,096: 2.55x, 3.68x, and 4.07x faster; allocations fell from 65,880 to 8,536, 600, and 197. - Run size 1: 10% slower with the same allocations. This is a partition extraction microbenchmark, not full writer throughput. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
