[ https://issues.apache.org/jira/browse/CASSANDRA-8160?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14329267#comment-14329267 ]
Benedict commented on CASSANDRA-8160: ------------------------------------- What's the rationale here? This only seems to make sense if the dataset is smaller than memory? And even then, using FADV_SEQUENTIAL seems unhelpful. All that offers is the OS the opportunity to _evict_ pages since it thinks we're done with them. WILLNEED asks it _not_ to, but it may do so at the expense of pages it really should retain. > CF level option to call posix_fadvise for sstables on creation and startup > -------------------------------------------------------------------------- > > Key: CASSANDRA-8160 > URL: https://issues.apache.org/jira/browse/CASSANDRA-8160 > Project: Cassandra > Issue Type: New Feature > Components: Core > Reporter: Matt Stump > Assignee: Branimir Lambov > Priority: Minor > Fix For: 2.1.4 > > Attachments: trunk-8160.txt > > > We should have a CF level configuration with will result in posix_fadvise > being called for sstables for that CF. It should be called on node startup > and for new sstables. This should be configurable per CF to allow for some > CFs to be prioritized above others. Not sure if we should use > POSIX_FADV_SEQUENTIAL or POSIX_FADV_WILLNEED. -- This message was sent by Atlassian JIRA (v6.3.4#6332)