[ https://issues.apache.org/jira/browse/CASSANDRA-9738?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14719307#comment-14719307 ]
Ariel Weisberg commented on CASSANDRA-9738: ------------------------------------------- For observers. We hit a snag. The key cache values can be quite large. There is an entry in each value for every row per partition so it can be in the thousands. This likely means that copying the entire thing on heap to operate on it once per read is not going to match the performance of the existing POJO implementation. Robert is going to benchmark a more representative configuration. It's a tractable problem but we will need to an off heap list implementation for variable size objects that supports binary search without materializing each entry in the search. > Migrate key-cache to be fully off-heap > -------------------------------------- > > Key: CASSANDRA-9738 > URL: https://issues.apache.org/jira/browse/CASSANDRA-9738 > Project: Cassandra > Issue Type: Sub-task > Reporter: Robert Stupp > Assignee: Robert Stupp > Fix For: 3.0 beta 2 > > > Key cache still uses a concurrent map on-heap. This could go to off-heap and > feels doable now after CASSANDRA-8099. > Evaluation should be done in advance based on a POC to prove that pure > off-heap counter cache buys a performance and/or gc-pressure improvement. > In theory, elimination of on-heap management of the map should buy us some > benefit. -- This message was sent by Atlassian JIRA (v6.3.4#6332)