[JBoss JIRA] (TEIID-2861) Partitioned UNION query fails
by Tom Arnold (JIRA)
Tom Arnold created TEIID-2861:
---------------------------------
Summary: Partitioned UNION query fails
Key: TEIID-2861
URL: https://issues.jboss.org/browse/TEIID-2861
Project: Teiid
Issue Type: Bug
Components: Query Engine
Affects Versions: 8.6
Reporter: Tom Arnold
Assignee: Steven Hawkins
I have a query like this where FOO and BAR are two different Oracle databases.
select * from (
(select item_id, created_at, 'Foo' source from foo.items order by created_at desc limit 5000)
union all
(select item_id, created_at, 'Bar' source from bar.items order by created_at desc limit 5000)
) x
where
source in ('Foo', 'Bar')
order by created_at desc
limit 0, 500;
If I adjust the filter to only include results from the second query, I get an exception.
09:15:19,287 ERROR [org.teiid.PROCESSOR] (Worker5_QueryProcessorQueue50) 7rl0TEA59SdU TEIID30019 Unexpected exception for request 7rl0TEA59SdU.7: java.lang.ArrayIndexOutOfBoundsException: -1
at java.util.ArrayList.elementData(ArrayList.java:400) [rt.jar:1.7.0_51]
at java.util.ArrayList.get(ArrayList.java:413) [rt.jar:1.7.0_51]
at org.teiid.query.optimizer.relational.rules.RuleRaiseNull.raiseNullNode(RuleRaiseNull.java:170) [teiid-engine-8.6.0.Final.jar:8.6.0.Final]
at org.teiid.query.optimizer.relational.rules.RuleRaiseNull.execute(RuleRaiseNull.java:70) [teiid-engine-8.6.0.Final.jar:8.6.0.Final]
at org.teiid.query.optimizer.relational.RelationalPlanner.executeRules(RelationalPlanner.java:606) [teiid-engine-8.6.0.Final.jar:8.6.0.Final]
at org.teiid.query.optimizer.relational.RelationalPlanner.optimize(RelationalPlanner.java:255) [teiid-engine-8.6.0.Final.jar:8.6.0.Final]
at org.teiid.query.optimizer.QueryOptimizer.optimizePlan(QueryOptimizer.java:159) [teiid-engine-8.6.0.Final.jar:8.6.0.Final]
at org.teiid.dqp.internal.process.Request.generatePlan(Request.java:408) [teiid-engine-8.6.0.Final.jar:8.6.0.Final]
at org.teiid.dqp.internal.process.Request.processRequest(Request.java:435) [teiid-engine-8.6.0.Final.jar:8.6.0.Final]
at org.teiid.dqp.internal.process.RequestWorkItem.processNew(RequestWorkItem.java:600) [teiid-engine-8.6.0.Final.jar:8.6.0.Final]
at org.teiid.dqp.internal.process.RequestWorkItem.process(RequestWorkItem.java:309) [teiid-engine-8.6.0.Final.jar:8.6.0.Final]
at org.teiid.dqp.internal.process.AbstractWorkItem.run(AbstractWorkItem.java:51) [teiid-engine-8.6.0.Final.jar:8.6.0.Final]
at org.teiid.dqp.internal.process.RequestWorkItem.run(RequestWorkItem.java:248) [teiid-engine-8.6.0.Final.jar:8.6.0.Final]
at org.teiid.dqp.internal.process.DQPWorkContext.runInContext(DQPWorkContext.java:273) [teiid-engine-8.6.0.Final.jar:8.6.0.Final]
at org.teiid.dqp.internal.process.ThreadReuseExecutor$RunnableWrapper.run(ThreadReuseExecutor.java:119) [teiid-engine-8.6.0.Final.jar:8.6.0.Final]
at org.teiid.dqp.internal.process.ThreadReuseExecutor$3.run(ThreadReuseExecutor.java:214) [teiid-engine-8.6.0.Final.jar:8.6.0.Final]
at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1145) [rt.jar:1.7.0_51]
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:615) [rt.jar:1.7.0_51]
at java.lang.Thread.run(Thread.java:744) [rt.jar:1.7.0_51]
--
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
10 years, 11 months
[JBoss JIRA] (TEIID-2860) Nested query with a with clause projection issue
by Steven Hawkins (JIRA)
Steven Hawkins created TEIID-2860:
-------------------------------------
Summary: Nested query with a with clause projection issue
Key: TEIID-2860
URL: https://issues.jboss.org/browse/TEIID-2860
Project: Teiid
Issue Type: Bug
Components: Query Engine
Affects Versions: 8.4
Reporter: Steven Hawkins
Assignee: Steven Hawkins
Priority: Critical
Fix For: 8.7
A view or inline view with a with clause is not projected properly, which can lead to serialization and other issues.
--
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
10 years, 11 months
[JBoss JIRA] (TEIID-2859) Provide identity generation for TEXTTABLE and OBJECTTABLE
by Ramesh Reddy (JIRA)
[ https://issues.jboss.org/browse/TEIID-2859?page=com.atlassian.jira.plugin... ]
Ramesh Reddy updated TEIID-2859:
--------------------------------
Forum Reference: https://community.jboss.org/thread/237251
> Provide identity generation for TEXTTABLE and OBJECTTABLE
> ---------------------------------------------------------
>
> Key: TEIID-2859
> URL: https://issues.jboss.org/browse/TEIID-2859
> Project: Teiid
> Issue Type: Enhancement
> Components: Query Engine
> Reporter: Ramesh Reddy
> Assignee: Steven Hawkins
>
> Extend the functionality FOR ORDINALITY defined on XMLTABLE, to TEXTTABLE and OBJECTTABLE.
> This will useful in providing a automatic identity key if the data is being read from sources where there is identity column provided. The faced the issue when loading a text file like
> {code}
> VA,F,1910,Mary,848
> VA,F,1911,Mary,747
> {code}
> Where name each line is a separate row, but they do not have unique identifier attached with them. Also helpful with interfaces like OData where they require a PK on a table. This provides an easier alternative.
--
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
10 years, 11 months
[JBoss JIRA] (TEIID-2859) Provide identity generation for TEXTTABLE and OBJECTTABLE
by Ramesh Reddy (JIRA)
Ramesh Reddy created TEIID-2859:
-----------------------------------
Summary: Provide identity generation for TEXTTABLE and OBJECTTABLE
Key: TEIID-2859
URL: https://issues.jboss.org/browse/TEIID-2859
Project: Teiid
Issue Type: Enhancement
Components: Query Engine
Reporter: Ramesh Reddy
Assignee: Steven Hawkins
Extend the functionality FOR ORDINALITY defined on XMLTABLE, to TEXTTABLE and OBJECTTABLE.
This will useful in providing a automatic identity key if the data is being read from sources where there is identity column provided. The faced the issue when loading a text file like
{code}
VA,F,1910,Mary,848
VA,F,1911,Mary,747
{code}
Where name each line is a separate row, but they do not have unique identifier attached with them. Also helpful with interfaces like OData where they require a PK on a table. This provides an easier alternative.
--
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
10 years, 11 months
[JBoss JIRA] (TEIID-2857) Build Error with Teiid POM Dependency
by Steven Hawkins (JIRA)
[ https://issues.jboss.org/browse/TEIID-2857?page=com.atlassian.jira.plugin... ]
Steven Hawkins updated TEIID-2857:
----------------------------------
Assignee: Van Halbert (was: Steven Hawkins)
So Ivan the resolution here for 8.7 will be to use a pre-release, but probably not a SNAPSHOT, for the bom. I'm not sure however that we can generally guarantee that something like a bom reference we don't control will not be a snapshot.
You'll also have to use an expanding settings.xml file (the developer setup page was corrected a while ago as it did reference a settings.xml that was not complete).
> Build Error with Teiid POM Dependency
> -------------------------------------
>
> Key: TEIID-2857
> URL: https://issues.jboss.org/browse/TEIID-2857
> Project: Teiid
> Issue Type: Bug
> Components: Embedded
> Affects Versions: 8.6
> Reporter: Ivan Chan
> Assignee: Van Halbert
> Labels: teiid
>
> In the jboss public maven repository:
> https://repository.jboss.org/nexus/content/groups/public
> The teiid-parent-8.6.0.Final.pom file has two dependencies which look to be
> Teiid build/configuration bugs.
> These two issues are causing our source code build to fail.
> The pom file is located here:
> https://repository.jboss.org/nexus/content/groups/public/org/jboss/teiid/...
> In this pom file there are two declared dependencies:
> <version.org.jboss.integration-platform>6.0.0-SNAPSHOT</version.org.jboss.integration-platform>
> <jbossas-version>7.2.0.Alpha1-redhat-4</jbossas-version>
> <dependency>
> <groupId>org.jboss.integration-platform</groupId>
> <artifactId>jboss-integration-platform-bom</artifactId>
> <version>${version.org.jboss.integration-platform}</version>
> <scope>import</scope>
> <type>pom</type>
> </dependency>
> <dependency>
> <groupId>org.jboss.as</groupId>
> <artifactId>jboss-as-parent</artifactId>
> <version>${jbossas-version}</version>
> <type>pom</type>
> <scope>import</scope>
> </dependency>
> Here are the issues:
> 1) Snapshot dependency existing in the 8.6.0-FINAL Release
> - jboss-integration-platform-bom/6.0.0-SNAPSHOT
> 2) Declared dependency does not exist in the repo repository.jboss.org
> - jboss-as-parent/7.2.0.Alpha1-redhat-4
> - We see dependencies only for:
> - 7.2.0.Alpha1-SNAPSHOT
> - 7.2.0.Fina
> I'm not sure why these dependencies are needed when pulling in the Teiid artifacts, but perhaps because the <scope> of
> these dependencies is set to "import" (<scope>import</scope>)
> Our usage of the repository.jboss.org is to have snapshots disabled. So that is why the 6.0.0-SNAPSHOT fails to be found.
> For the 7.2.0.Alpha1-redhat-4 dependency, it does not exist in
> repository.jboss.org so this also fails.
--
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
10 years, 11 months
[JBoss JIRA] (TEIID-2858) Database use of char[>1] cause using Convert(x, x) and negates using an index
by Steven Hawkins (JIRA)
[ https://issues.jboss.org/browse/TEIID-2858?page=com.atlassian.jira.plugin... ]
Steven Hawkins commented on TEIID-2858:
---------------------------------------
I still don't think we're on the same page. You'll need to be quite explicit as to what is happening. If you have a char[x] modeled as string, there should be no convert involved. Can you provide a scenario where this is happening?
Yes, Tom that is understood. As an aside we do now have limited support for function based indexes on internal materialized views.
> Database use of char[>1] cause using Convert(x,x) and negates using an index
> ----------------------------------------------------------------------------
>
> Key: TEIID-2858
> URL: https://issues.jboss.org/browse/TEIID-2858
> Project: Teiid
> Issue Type: Bug
> Components: Misc. Connectors
> Affects Versions: 8.4.1
> Reporter: Van Halbert
> Assignee: Steven Hawkins
>
> When a database defines a char of length greater than 1, a convert(x,x) is being used in the criteria, and thereby, negating the use of an index and causing table scans.
> Note: There is no option to change the database schemas, because they are from a COTS product.
> I would think there needs to be a solution in the translator such that if it can be detected that its for a char column and that it does not do the convert.
--
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
10 years, 11 months
[JBoss JIRA] (TEIID-2858) Database use of char[>1] cause using Convert(x, x) and negates using an index
by Tom Johnston (JIRA)
[ https://issues.jboss.org/browse/TEIID-2858?page=com.atlassian.jira.plugin... ]
Tom Johnston commented on TEIID-2858:
-------------------------------------
As a general rule any function will negate the use of any index on the field. Though there are "function-based-indexes" these are not options with COTS products. Some consideration must be made when a potential fix has ramifications that would diminish the performance. "Middleware" already has something to prove when it comes to NOT heavily impacting performance so let's not purposely do something that can impact performance.
> Database use of char[>1] cause using Convert(x,x) and negates using an index
> ----------------------------------------------------------------------------
>
> Key: TEIID-2858
> URL: https://issues.jboss.org/browse/TEIID-2858
> Project: Teiid
> Issue Type: Bug
> Components: Misc. Connectors
> Affects Versions: 8.4.1
> Reporter: Van Halbert
> Assignee: Steven Hawkins
>
> When a database defines a char of length greater than 1, a convert(x,x) is being used in the criteria, and thereby, negating the use of an index and causing table scans.
> Note: There is no option to change the database schemas, because they are from a COTS product.
> I would think there needs to be a solution in the translator such that if it can be detected that its for a char column and that it does not do the convert.
--
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
10 years, 11 months