falcon-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Srikanth Sundarrajan (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (FALCON-141) Support cluster updates
Date Thu, 10 Dec 2015 06:55:11 GMT

    [ https://issues.apache.org/jira/browse/FALCON-141?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15050208#comment-15050208
] 

Srikanth Sundarrajan commented on FALCON-141:
---------------------------------------------

[~bvellanki], When you update a feed, all process need to be updated, If any process update
were to fail, we don't flag that as pending update and have any off-band mechanism to finish
the pending update. Entity updates are idempotent and can repeatedly attempted till it succeeds
without any side effect. Is there a reason for cluster update to behave differently ?

May be we can add a note to say that older instances need special handling, but from our own
experiences, they have been painful to manage. Another issue that we might encounter during
a cluster end point update is when instances are materialized in Oozie (which are not in waiting
state). These ones will invariably fail post update and might require hand holding.

> Support cluster updates
> -----------------------
>
>                 Key: FALCON-141
>                 URL: https://issues.apache.org/jira/browse/FALCON-141
>             Project: Falcon
>          Issue Type: Bug
>            Reporter: Shwetha G S
>            Assignee: Balu Vellanki
>




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

Mime
View raw message