[ https://issues.apache.org/jira/browse/CASSANDRA-8714?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14338854#comment-14338854 ]
Ariel Weisberg commented on CASSANDRA-8714: ------------------------------------------- I am +1 with one comment. Thinking on the change to MemoryUtil.setBytes, you switched to getting the address by casting to an interface and then invoking vs accessing the field by offset. setBytes might loop over a fair number of bytes so startup overhead is not as important, but it could also be used to copy just a few so startup time will matter there. If the JVM can figure out that the ByteBuffer in play there is always the same kind (or maybe the method is always the same one) it will de-virtualize and inline, but testing for that is beyond me. [~benedict] might know what will happen. > row-cache: use preloaded jemalloc w/ Unsafe > ------------------------------------------- > > Key: CASSANDRA-8714 > URL: https://issues.apache.org/jira/browse/CASSANDRA-8714 > Project: Cassandra > Issue Type: Sub-task > Reporter: Robert Stupp > Assignee: Robert Stupp > Fix For: 3.0 > > Attachments: 8714-2.txt, 8714-3.txt, 8714-4.txt, 8714.txt > > > Using jemalloc via Java's {{Unsafe}} is a better alternative on Linux than > using jemalloc via JNA. -- This message was sent by Atlassian JIRA (v6.3.4#6332)