[ https://issues.apache.org/jira/browse/CASSANDRA-9830?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15218622#comment-15218622 ]
Marcus Eriksson commented on CASSANDRA-9830: -------------------------------------------- bq. The basic idea is to regenerate BF if dropping to level 0 during anti-compaction Problem is that if the repaired sstable fits within the repaired range we don't anticompact, we just mutate the repairedAt value for the sstable [here|https://github.com/apache/cassandra/blob/trunk/src/java/org/apache/cassandra/db/compaction/CompactionManager.java#L610-L621] > Option to disable bloom filter in highest level of LCS sstables > --------------------------------------------------------------- > > Key: CASSANDRA-9830 > URL: https://issues.apache.org/jira/browse/CASSANDRA-9830 > Project: Cassandra > Issue Type: New Feature > Components: Compaction > Reporter: Jonathan Ellis > Assignee: Paulo Motta > Priority: Minor > Labels: performance > Fix For: 3.x > > > We expect about 90% of data to be in the highest level of LCS in a fully > populated series. (See also CASSANDRA-9829.) > Thus if the user is primarily asking for data (partitions) that has actually > been inserted, the bloom filter on the highest level only helps reject > sstables about 10% of the time. > We should add an option that suppresses bloom filter creation on top-level > sstables. This will dramatically reduce memory usage for LCS and may even > improve performance as we no longer check a low-value filter. > (This is also an idea from RocksDB.) -- This message was sent by Atlassian JIRA (v6.3.4#6332)