[
https://issues.apache.org/jira/browse/KUDU-2017?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16013289#comment-16013289
]
Todd Lipcon commented on KUDU-2017:
-----------------------------------
Credit to [~adar] and [~danburkert] for finding this
> Incorrect calculation determining performance improvement for flush
> -------------------------------------------------------------------
>
> Key: KUDU-2017
> URL: https://issues.apache.org/jira/browse/KUDU-2017
> Project: Kudu
> Issue Type: Bug
> Components: tablet
> Affects Versions: 1.3.1
> Reporter: Todd Lipcon
> Assignee: Todd Lipcon
>
> The code to calculate the MM "perf improvement score" for a flush had two
> bugs:
> (1) we calculated {{threshold - current_usage}} instead of {{current_usage -
> threshold}}, which resulted in a negative score
> (2) we had an unsigned integer overflow, which resulted in the above negative
> score becoming insanely large
> These two wrongs "made a right" in which we'd still trigger flushes at the
> flush threshold, which is why this went unnoticed for quite some time.
> However, the flushing behavior is more aggressive than originally intended,
> and we would lose the correct prioritization of flushing tablets that are
> farther above the threshold.
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
