[jira] [Commented] (ACCUMULO-3845) DurabilityIT failed

["Josh Elser (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/ACCUMULO-3845?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14555484#comment-14555484 ] 

Josh Elser commented on ACCUMULO-3845:
--------------------------------------

Looks like median is a good choice (sync, flush, log, none: 5x each):

{noformat}
2015-05-21 22:39:39,706 [functional.DurabilityIT] INFO : Attempt durations: [676, 707, 737, 825, 1582]
2015-05-21 22:39:43,213 [functional.DurabilityIT] INFO : Attempt durations: [577, 582, 623, 666, 728]
2015-05-21 22:39:46,570 [functional.DurabilityIT] INFO : Attempt durations: [556, 580, 590, 608, 673]
2015-05-21 22:39:49,379 [functional.DurabilityIT] INFO : Attempt durations: [425, 433, 454, 481, 739]
{noformat}

> DurabilityIT failed
> -------------------
>
>                 Key: ACCUMULO-3845
>                 URL: https://issues.apache.org/jira/browse/ACCUMULO-3845
>             Project: Accumulo
>          Issue Type: Bug
>          Components: test
>            Reporter: Josh Elser
>            Assignee: Josh Elser
>            Priority: Minor
>             Fix For: 1.8.0, 1.7.1
>
>
> {noformat}
> sync 1933 flush 1313 log 1420 none 991
> {noformat}
> Not anything wrong, just sensitive. We could run the write tests a few times and take the median. Hopefully that will even things out.
> The assertion message "sync should be faster than log" is also wrong :)



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