[jira] [Commented] (CASSANDRA-12580) Fix merkle tree size calculation

["Yuki Morishita (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/CASSANDRA-12580?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15528174#comment-15528174 ] 

Yuki Morishita commented on CASSANDRA-12580:
--------------------------------------------

Nice catch. Patch looks good to me.


> Fix merkle tree size calculation
> --------------------------------
>
>                 Key: CASSANDRA-12580
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-12580
>             Project: Cassandra
>          Issue Type: Bug
>            Reporter: Paulo Motta
>            Assignee: Paulo Motta
>             Fix For: 2.1.x, 2.2.x, 3.0.x, 3.x
>
>
> On CASSANDRA-5263 it was introduced dynamic merkle tree sizing based on estimated number of partitions as {{estimatedDepth = lg(numPartitions)}}, but on [CompactionManager.doValidationCompaction|https://github.com/apache/cassandra/blob/cassandra-2.1/src/java/org/apache/cassandra/db/compaction/CompactionManager.java#L1052] this is being calculated as:
> {{int depth = numPartitions > 0 ? (int) Math.min(Math.floor(Math.log(numPartitions)), 20) : 0;}}
> This is actually calculating {{ln(numPartitions)}} (base-e) instead of {{lg(numPartitions)}} (base-2), which causes merkle trees to lose resolution, what may result in overstreaming.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)