You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@falcon.apache.org by "sandeep samudrala (JIRA)" <ji...@apache.org> on 2015/09/03 09:08:46 UTC

[jira] [Updated] (FALCON-298) Feed update with replication delay

     [ https://issues.apache.org/jira/browse/FALCON-298?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

sandeep samudrala updated FALCON-298:
-------------------------------------
    Attachment: FALCON-298.patch

The delays can be different for each source cluster and there by figure out the source cluster, which the current coordinator belongs to and then get the delay and then add it to the end time of the current old coordinator in the target.

> Feed update with replication delay
> ----------------------------------
>
>                 Key: FALCON-298
>                 URL: https://issues.apache.org/jira/browse/FALCON-298
>             Project: Falcon
>          Issue Type: Bug
>            Reporter: Shwetha G S
>            Assignee: sandeep samudrala
>              Labels: newbie
>         Attachments: FALCON-298.patch
>
>
> When feed is updated, falcon sets endtime of old coordinator to now and creates new coord with start time as now. But for replication update with delay, old coord's endtime is set to now and new coord's start time is set to now+delay. This creates a hole in scheduler



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