mirror of
https://github.com/scylladb/scylladb.git
synced 2026-04-27 11:55:15 +00:00
Main assumption here is that if is_big is good enough for GetBatchItems operation it should work well also for Scan, Query and GetRecords. And it's easier to maintain more unified code. Additionally 'future<> print' documentation used for streaming suggests that there is quite big overhead so since it seems the only motivation for streaming was to reduce contiguous allocation size below some threshold we should not stream when this threshold is not exceeded. Closes #12164
223 KiB
223 KiB