[
https://issues.apache.org/jira/browse/FALCON-96?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13751677#comment-13751677
]
Venkatesh Seetharam commented on FALCON-96:
-------------------------------------------
bq. When you say RPC using thrift are you thinking of talking to HiveServer2 or
directly to the thrift metastore server?
I'm using org.apache.hadoop.hive.metastore.HiveMetaStoreClient which is not
HiveServer2 I suppose.
I agree that using webhcat makes total sense but webhcat using hcat cli is very
bad. Falcon brings up all services during integration tests and having to have
hcat install is not viable.
I'm leaning towards using HiveMetaStoreClient and eventually migrating it to
webhcat.
> Hive client to talk to the metastore
> ------------------------------------
>
> Key: FALCON-96
> URL: https://issues.apache.org/jira/browse/FALCON-96
> Project: Falcon
> Issue Type: Sub-task
> Affects Versions: 0.3
> Reporter: Venkatesh Seetharam
>
> There are 2 choices: RPC using thrift and HTTP using webhcat. HTTP of course
> is preferable but there are a few issues with it:
> * Uses cli to talk to the server implying needing Hcat install and client.
> This is a PITA for integration-tests.
> * Also, there is no java wrapper around the RESTful APIs which returns Map
> and such. Need to roll our own.
> * Runs as a separate daemon (which should eventually fold into metastore or
> HiveServer2). However this is short-term and a minor annoyance.
> Thoughts?
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
