[
https://issues.apache.org/jira/browse/CASSANDRA-2062?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12987772#action_12987772
]
Stu Hood commented on CASSANDRA-2062:
-------------------------------------
johano pointed out that http://code.google.com/p/caliper/ would be a good way
to confirm that this doesn't break our common case usages of Collating/Merge.
I'm going to perform the Reducing+MergeIterator smash tonight, and then I'll
probably spend microtime microbenchmarking.
> Use more efficient merge algorithm
> ----------------------------------
>
> Key: CASSANDRA-2062
> URL: https://issues.apache.org/jira/browse/CASSANDRA-2062
> Project: Cassandra
> Issue Type: Improvement
> Components: Core
> Reporter: Stu Hood
> Priority: Minor
> Fix For: 0.7.2
>
> Attachments: 0001-Improved-iterator-for-merging-sorted-iterators.txt,
> 0002-Quickie-instrumentation-for-comparisons.txt,
> 0003-Replace-Collating-with-Merge-in-CompactionIterator.txt,
> 0004-Port-LazilyCompactedRow-ReducingKeyIterator-RangeSlice.txt,
> 0005-Remove-temporary-instrumentation-and-CollatingIterator.txt
>
>
> For {{M}} iterators containing {{N}} total items,
> commons.collections.CollatingIterator performs a {{O(M*N)}} merge, and calls
> hasNext multiple times per returned value. We can do better.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
