You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@falcon.apache.org by "Sowmya Ramesh (JIRA)" <ji...@apache.org> on 2015/11/06 20:51:10 UTC
[jira] [Updated] (FALCON-1569) Bug in setting the frequency of Feed
retention coordinator
[ https://issues.apache.org/jira/browse/FALCON-1569?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Sowmya Ramesh updated FALCON-1569:
----------------------------------
Attachment: FALCON-1569.V0.patch
> Bug in setting the frequency of Feed retention coordinator
> -----------------------------------------------------------
>
> Key: FALCON-1569
> URL: https://issues.apache.org/jira/browse/FALCON-1569
> Project: Falcon
> Issue Type: Bug
> Components: retention
> Reporter: Sowmya Ramesh
> Assignee: Sowmya Ramesh
> Fix For: trunk
>
> Attachments: FALCON-1569.V0.patch
>
>
> Currently in FeedRetentionCoordinatorBuilder, timeUnit is used to determine frequency of the retention coordinator.
> {code}
> TimeUnit timeUnit = entity.getFrequency().getTimeUnit();
> if (timeUnit == TimeUnit.hours || timeUnit == TimeUnit.minutes) {
> coord.setFrequency("${coord:hours(6)}");
> } else {
> coord.setFrequency("${coord:days(1)}");
> }
> {code}
> days(2) can be mapped to hours(48). If user uses hours(48) then retention coordinator runs every 6 hours instead of running daily wasting the compute resources. Instead get the time duration and use that to determine the retention job frequency.
> Also fix it in FalconUnitTestBase.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)