[
https://issues.apache.org/jira/browse/TEZ-3404?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15471906#comment-15471906
]
Hitesh Shah commented on TEZ-3404:
----------------------------------
Comments ( mostly minor - thanks for the patience in addressing comments - I
think this will be the last round) :
{code}
TezConfiguration tezConf = new
TezConfiguration(amConfig.getTezConfiguration());
// Merge the dag access controls into tez am config.
if (dag != null && dag.getDagAccessControls() != null) {
dag.getDagAccessControls().mergeIntoAmAcls(tezConf);
}
{code}
- the new conf should only be cloned if required i.e. should be inside if
clause. Doc mentions what is being done but would be good to clarify why this
is being done.
{code}
// Fallback to session domain, if domainInfo was null or there was an
exception.
return sessionDomainId;
{code}
- the stmt seems a bit incorrect as this is not in a finally block. Also,
why return the sessionDomainId when as the dag is being added to the skipped
list on an exception. Seems like there should be a return null stmt in
exception catch block?
It seems like createSessionDomain and createDagDomain could be moved into a
single code location somewhere under the tez-yarn-timeline-history module and
invoked from each plugin impl as needed? Not needed in this jira but something
to consider as part of the larger re-factor.
> Remove blocking call for ATS domain creation from client side to AM
> -------------------------------------------------------------------
>
> Key: TEZ-3404
> URL: https://issues.apache.org/jira/browse/TEZ-3404
> Project: Apache Tez
> Issue Type: Improvement
> Reporter: Harish Jaiprakash
> Assignee: Harish Jaiprakash
> Attachments: TEZ-3404-09.patch, TEZ-3404-10.patch,
> TEZ-3404-WIP-01.patch, TEZ-3404-WIP-02.patch, TEZ-3404-WIP-03.patch,
> TEZ-3404-WIP-04.patch, TEZ-3404-WIP-05.patch, TEZ-3404-WIP-06.patch,
> TEZ-3404-WIP-07.patch, TEZ-3404-WIP-08.patch
>
>
> Today, there is a blocking call for creating the domain on the client side
> which blocks hive query throughput. We need to move this to the AM and still
> retain the same security guarantees.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
