[
https://issues.apache.org/jira/browse/HIVE-8225?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Sergey Shelukhin updated HIVE-8225:
-----------------------------------
Attachment: HIVE-8225.inprogress.patch
Here's partial fix in process... first I tried using Optiq rules but that
doesn't work due to some problems with transform (described in the patch).
I added a fix to post-CBO plan processing (before AST conversion), which fixes
the issue for one small query that originally had the problem (see q file), but
not for the original query. I'd need to take a look at that (probably the fix
needs to be in a better place, or some check mismatches), for now attaching the
in-progress patch since I will be on vacation soon
> CBO trunk merge: union11 test fails due to incorrect plan
> ---------------------------------------------------------
>
> Key: HIVE-8225
> URL: https://issues.apache.org/jira/browse/HIVE-8225
> Project: Hive
> Issue Type: Bug
> Reporter: Sergey Shelukhin
> Assignee: Sergey Shelukhin
> Attachments: HIVE-8225.inprogress.patch
>
>
> The result changes to as if the union didn't have count() inside. The issue
> can be fixed by using srcunion.value outside the subquery in count (replace
> count(1) with count(srcunion.value)). Otherwise, it looks like count(1) node
> from union-ed queries is not present in AST at all, which might cause this
> result.
> -Interestingly, adding group by to each query in a union produces completely
> weird result (count(1) is 309 for each key, whereas it should be 1 and the
> "logical" incorrect value if internal count is lost is 500)- Nm, that groups
> by table column called key, which is weird but is what Hive does
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
