Github user aweisberg commented on a diff in the pull request:
https://github.com/apache/cassandra/pull/239#discussion_r202486275
--- Diff: conf/cassandra.yaml ---
@@ -784,6 +784,15 @@ compaction_throughput_mb_per_sec: 16
# between the sstables, reducing page cache churn and keeping hot rows hot
sstable_preemptive_open_interval_in_mb: 50
+# When enabled, permits Cassandra to zero copy eligible SSTables between
--- End diff --
Worth documenting this optimization relies on having sstables that are
wholly contained in the range being streamed and that configurations like size
tiered without 6696 won't benefit, and would in fact spend time reading indexes
checking to see if the optimization applies? Leveled should always benefit from
this optimization.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
