[
https://issues.apache.org/jira/browse/HAWQ-10?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14958166#comment-14958166
]
ASF GitHub Bot commented on HAWQ-10:
------------------------------------
Github user liming01 commented on the pull request:
https://github.com/apache/incubator-hawq/pull/13#issuecomment-148253345
@jimmida
I haven't started to think about 'hcatalog' feature's impact for
gpmigrator. Some points need to be thinked:
1) If old db instance already has db named 'hcatalog': maybe we should
rename it?
2) Add a new meta for 'hcatalog' in pg_database: Done.
3) Skipping 'hcatalog' for any db loop processing. Because this affected
by the solution for point 1), so I haven't fixed it. However this behavior
works fine in this version's upgrade, it will have problem when next version's
upgrade.
But for this defect, this fix is okay, because the function
PerformPostUpgrade() in gpMigratorUtil.py calls gp_toolkit.sql.
> hcatalog database causes errors for hawq_toolkit.hawq_size_of_database
> ----------------------------------------------------------------------
>
> Key: HAWQ-10
> URL: https://issues.apache.org/jira/browse/HAWQ-10
> Project: Apache HAWQ
> Issue Type: Bug
> Components: PXF
> Reporter: Lirong Jian
> Assignee: Lirong Jian
> Original Estimate: 24h
> Remaining Estimate: 24h
>
> Since the hcatalog database is not a normal database, it would cause some
> errors when querying hawq_toolkit.hawq_size_of_database.
> Following is an example:
> SELECT * FROM hawq_toolkit.hawq_size_of_database;
> ERROR: database with OID 6120 does not exist (dbsize.c:219)
> OID 6120 database is the hcatalog database.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
