Home | About | Sematext search-lucene.com search-hadoop.com
NEW: Monitor These Apps!
elasticsearch, apache solr, apache hbase, hadoop, redis, casssandra, amazon cloudwatch, mysql, memcached, apache kafka, apache zookeeper, apache storm, ubuntu, centOS, red hat, debian, puppet labs, java, senseiDB
 Search Hadoop and all its subprojects:

Switch to Plain View
Hive >> mail # dev >> Re: [jira] [Commented] (HIVE-3999) Mysql metastore upgrade script will end up with different schema than the full schema load


Copy link to this message
-
Re: [jira] [Commented] (HIVE-3999) Mysql metastore upgrade script will end up with different schema than the full schema load
Hi Jarek Jarcec Cecho,
I agree with you now. Your solution is the most suitable way.
2013/2/18 Jarek Jarcec Cecho (JIRA) <[EMAIL PROTECTED]>

>
>     [
> https://issues.apache.org/jira/browse/HIVE-3999?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13580457#comment-13580457]
>
> Jarek Jarcec Cecho commented on HIVE-3999:
> ------------------------------------------
>
> Hi sir,
> setting the default value was one of the options that I've explored when
> working on this JIRA. However I've decided not to use it and rather use
> three step column addition (please checkout the patch). One of the reasons
> behind this decision is that with no explicit default value, running old
> hive against new metastore will fail. I believe that this is much safer as
> otherwise Hive would continue to operate on updated schema repository and
> might eventually corrupt the data.
>
> Jarcec
>
> > Mysql metastore upgrade script will end up with different schema than
> the full schema load
> >
> ------------------------------------------------------------------------------------------
> >
> >                 Key: HIVE-3999
> >                 URL: https://issues.apache.org/jira/browse/HIVE-3999
> >             Project: Hive
> >          Issue Type: Bug
> >          Components: Metastore
> >            Reporter: Jarek Jarcec Cecho
> >            Assignee: Jarek Jarcec Cecho
> >             Fix For: 0.11.0
> >
> >         Attachments: mysql_upgrade_issue.patch
> >
> >
> > I've noticed that the file {{hive-schema-0.10.0.mysql.sql}} is creating
> table SDS with following column:
> > {code}
> >   `IS_STOREDASSUBDIRECTORIES` bit(1) NOT NULL,
> > {code}
> > However the upgrade script {{011-HIVE-3649.mysql.sql}} will create the
> column differently:
> > {code}
> > ALTER TABLE `SDS` ADD `IS_STOREDASSUBDIRECTORIES` bit(1) ;
> > {code}
> > Thus user will get slightly different schema each time - once with "NOT
> NULL" and secondly with "NULL" definition.
>
> --
> 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
>

--
Best wishs´╝ü
Fangkun.Cao
NEW: Monitor These Apps!
elasticsearch, apache solr, apache hbase, hadoop, redis, casssandra, amazon cloudwatch, mysql, memcached, apache kafka, apache zookeeper, apache storm, ubuntu, centOS, red hat, debian, puppet labs, java, senseiDB