[
https://issues.apache.org/jira/browse/FALCON-141?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15049811#comment-15049811
]
Venkat Ranganathan commented on FALCON-141:
-------------------------------------------
[~sriksun] Thanks for the update.
bq. Would this be necessary, as all entities that are dependent on the cluster
should go through an update. Or in other words requireUpdate is derivable and
it may be unnecessary to maintain this as an explicit state.
I think we need to have a mechanism to say what have been updated post cluster
update so that failed updates can be re-attempted later - not to identify what
to change after the cluster update work. as you said
bq. Also what happens for entity instances which were complete before the
cluster update and which may require a re-run after the update
Failed instances before the update with old coordinator will not run - One
way is to create the new coordinators only when JT/RM or NN interfaces change
and otherwise update existing entities so that reruns are handled correctly in
those cases. And whenever JT/RM or NN endpoints change, we can issue a
warning saying that post the update and procesing, reruns for instances prior
to the update cannot be done
> 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)
