[JBoss JIRA] Commented: (TEIID-1651) Due to a change in the JDBC spec on how the use of "AS" in the select statement will now be returned in the getColumnLabel of the ResultSetMetadata, Teiid must provide backwards support
by Steven Hawkins (JIRA)
[ https://issues.jboss.org/browse/TEIID-1651?page=com.atlassian.jira.plugin... ]
Steven Hawkins commented on TEIID-1651:
---------------------------------------
The default behavior for JDBC 4.0 should be true, not false. The only reason to default to false would be to match the EDS 5.1 behavior, but that would actually be preserving a bug. I would vote for defaulting to true with an entry in the release notes on setting to false. Also where do you see that useJDBC4ColumnNameAndLabelSemantics is a spec property? It seems that property key is just something used by db2 and related such as derby. For example mysql uses useOldAliasMetadataBehavior.
Comparing to the db2 docs our old findColumn logic was based upon only column labels (which means that it wasn't quite inline with the spec), so there is no behavioral change there.
> Due to a change in the JDBC spec on how the use of "AS" in the select statement will now be returned in the getColumnLabel of the ResultSetMetadata, Teiid must provide backwards support
> ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
>
> Key: TEIID-1651
> URL: https://issues.jboss.org/browse/TEIID-1651
> Project: Teiid
> Issue Type: Bug
> Components: JDBC Driver
> Affects Versions: 7.4.1
> Reporter: Van Halbert
> Assignee: Steven Hawkins
> Priority: Blocker
>
> Teiid has already applied a fix to support the JDBC spec change for when an alias is used in the select clause. The spec change says the "AS" name will now be returned when getColumnLabel() is called, instead of getColumnName(). Unfortunately, this breaks backwards compatibility. To resolve the issue, the JDBC driver, to provide support for backwards compatibility, can allow the user to specify the useJDBC4ColumnNameAndLabelSemantics=true (default is false) property on the url (this is the recommended spec property).
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
14 years, 9 months
[JBoss JIRA] Updated: (TEIID-1651) Due to a change in the JDBC spec on how the use of "AS" in the select statement will now be returned in the getColumnLabel of the ResultSetMetadata, Teiid must provide backwards support
by Van Halbert (JIRA)
[ https://issues.jboss.org/browse/TEIID-1651?page=com.atlassian.jira.plugin... ]
Van Halbert updated TEIID-1651:
-------------------------------
Description: Teiid has already applied a fix to support the JDBC spec change for when an alias is used in the select clause. The spec change says the "AS" name will now be returned when getColumnLabel() is called, instead of getColumnName(). Unfortunately, this breaks backwards compatibility. To resolve the issue, the JDBC driver, to provide support for backwards compatibility, can allow the user to specify the useJDBC4ColumnNameAndLabelSemantics=true (default is false) property on the url (this is the recommended spec property). (was: Teiid has already applied a fix to support the JDBC spec change for when an alias is used in the select clause. The spec change says the "AS" name will now be returned when getColumnLabel() is called, instead of getColumnName(). Unfortunately, this breaks backwards compatibility in 5.2 and any vendor/client that is using EDS and depending on metadata. To resolve the issue, the JDBC driver, to provide support for backwards compatibility, can allow the user to specify the useJDBC4ColumnNameAndLabelSemantics=true (default is false) property on the url (this is the recommended spec property). Therefore, for 5.2, the default for this property should be set to "TRUE", so that the user doesn't have to make any changes in order to expect the same behavior. I would recommend in a future major release (i.e., 6.0) this default be changed to the industry standard of "TRUE".)
> Due to a change in the JDBC spec on how the use of "AS" in the select statement will now be returned in the getColumnLabel of the ResultSetMetadata, Teiid must provide backwards support
> ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
>
> Key: TEIID-1651
> URL: https://issues.jboss.org/browse/TEIID-1651
> Project: Teiid
> Issue Type: Bug
> Components: JDBC Driver
> Affects Versions: 7.4.1
> Reporter: Van Halbert
> Assignee: Steven Hawkins
> Priority: Blocker
>
> Teiid has already applied a fix to support the JDBC spec change for when an alias is used in the select clause. The spec change says the "AS" name will now be returned when getColumnLabel() is called, instead of getColumnName(). Unfortunately, this breaks backwards compatibility. To resolve the issue, the JDBC driver, to provide support for backwards compatibility, can allow the user to specify the useJDBC4ColumnNameAndLabelSemantics=true (default is false) property on the url (this is the recommended spec property).
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
14 years, 9 months
[JBoss JIRA] Assigned: (TEIID-1651) Due to a change in the JDBC spec on how the use of "AS" in the select statement will now be returned in the getColumnLabel of the ResultSetMetadata, Teiid must provide backwards support
by Van Halbert (JIRA)
[ https://issues.jboss.org/browse/TEIID-1651?page=com.atlassian.jira.plugin... ]
Van Halbert reassigned TEIID-1651:
----------------------------------
Assignee: Steven Hawkins (was: Van Halbert)
> Due to a change in the JDBC spec on how the use of "AS" in the select statement will now be returned in the getColumnLabel of the ResultSetMetadata, Teiid must provide backwards support
> ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
>
> Key: TEIID-1651
> URL: https://issues.jboss.org/browse/TEIID-1651
> Project: Teiid
> Issue Type: Bug
> Components: JDBC Driver
> Affects Versions: 7.4.1
> Reporter: Van Halbert
> Assignee: Steven Hawkins
> Priority: Blocker
>
> Teiid has already applied a fix to support the JDBC spec change for when an alias is used in the select clause. The spec change says the "AS" name will now be returned when getColumnLabel() is called, instead of getColumnName(). Unfortunately, this breaks backwards compatibility in 5.2 and any vendor/client that is using EDS and depending on metadata. To resolve the issue, the JDBC driver, to provide support for backwards compatibility, can allow the user to specify the useJDBC4ColumnNameAndLabelSemantics=true (default is false) property on the url (this is the recommended spec property). Therefore, for 5.2, the default for this property should be set to "TRUE", so that the user doesn't have to make any changes in order to expect the same behavior. I would recommend in a future major release (i.e., 6.0) this default be changed to the industry standard of "TRUE".
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
14 years, 9 months
[JBoss JIRA] Moved: (TEIID-1651) Due to a change in the JDBC spec on how the use of "AS" in the select statement will now be returned in the getColumnLabel of the ResultSetMetadata, Teiid must provide backwards support
by Van Halbert (JIRA)
[ https://issues.jboss.org/browse/TEIID-1651?page=com.atlassian.jira.plugin... ]
Van Halbert moved SOA-3121 to TEIID-1651:
-----------------------------------------
Project: Teiid (was: JBoss Enterprise SOA Platform)
Key: TEIID-1651 (was: SOA-3121)
Affects Version/s: 7.4.1
(was: 5.2.0 GA)
Component/s: JDBC Driver
(was: EDS)
Security: (was: JBoss Internal)
> Due to a change in the JDBC spec on how the use of "AS" in the select statement will now be returned in the getColumnLabel of the ResultSetMetadata, Teiid must provide backwards support
> ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
>
> Key: TEIID-1651
> URL: https://issues.jboss.org/browse/TEIID-1651
> Project: Teiid
> Issue Type: Bug
> Components: JDBC Driver
> Affects Versions: 7.4.1
> Reporter: Van Halbert
> Assignee: Van Halbert
> Priority: Blocker
>
> Teiid has already applied a fix to support the JDBC spec change for when an alias is used in the select clause. The spec change says the "AS" name will now be returned when getColumnLabel() is called, instead of getColumnName(). Unfortunately, this breaks backwards compatibility in 5.2 and any vendor/client that is using EDS and depending on metadata. To resolve the issue, the JDBC driver, to provide support for backwards compatibility, can allow the user to specify the useJDBC4ColumnNameAndLabelSemantics=true (default is false) property on the url (this is the recommended spec property). Therefore, for 5.2, the default for this property should be set to "TRUE", so that the user doesn't have to make any changes in order to expect the same behavior. I would recommend in a future major release (i.e., 6.0) this default be changed to the industry standard of "TRUE".
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
14 years, 9 months
[JBoss JIRA] Resolved: (TEIID-1638) Add ability to query parent record info from child table query without joins for the Salesforce Translator
by Steven Hawkins (JIRA)
[ https://issues.jboss.org/browse/TEIID-1638?page=com.atlassian.jira.plugin... ]
Steven Hawkins resolved TEIID-1638.
-----------------------------------
Resolution: Deferred
There are no plans at this time to directly support SOQL semantics. Outer join based queries should be eligible for pushdown. We will also at some point provide encapsulated native query support. Those are the recommended approaches.
> Add ability to query parent record info from child table query without joins for the Salesforce Translator
> ----------------------------------------------------------------------------------------------------------
>
> Key: TEIID-1638
> URL: https://issues.jboss.org/browse/TEIID-1638
> Project: Teiid
> Issue Type: Feature Request
> Components: Salesforce Connector
> Affects Versions: 7.1.1
> Reporter: Debbie Steigner
> Labels: eds, salesforce
>
> Salesforce.com SOQL supports selecting parent record value through reference key name.
> We need to query parent record info from child table query. For example, if I have column parentid in Contact object which is reference key to account object, I can do in SOQL in salesforce.com something like this.
> Select c.Id, c.Name, c.AccountId, c.Account.Name
> from Contact c
> But if i want to do the same in EDS, I have to use Relationship Queries (join of two tables Account and Contact) which is huge impact on performance
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
14 years, 9 months
[JBoss JIRA] Created: (TEIID-1649) Reading CLOB's from Teradata has issues and produces exception when stream is read from a different thread than what the clob was read from the resultset
by Van Halbert (JIRA)
Reading CLOB's from Teradata has issues and produces exception when stream is read from a different thread than what the clob was read from the resultset
---------------------------------------------------------------------------------------------------------------------------------------------------------
Key: TEIID-1649
URL: https://issues.jboss.org/browse/TEIID-1649
Project: Teiid
Issue Type: Bug
Components: Misc. Connectors
Affects Versions: 7.4.1
Reporter: Van Halbert
Assignee: Steven Hawkins
Priority: Critical
The initial problem was seen when running BQT testing against the teradata source. However, when debugging, the exception was occurring on the initial client read of the stream. After reading teradata jdbc driver doc, it talks about how its not thread safe and has issues when different threads are handling blobs/clobs. Which lead me to do a simple client test. 2 tests where run run thru a servlet that obtains the teradata connenction via jndi. Thet are as follows (attached is the sample code after the changes used for #2):
1. simple execute/select: execute select intkey, objectvalue from bqt1.smalla, for each row, obtain the clob and read the stream (no exception is seen)
2. execute select and read clob in a different thread: a new thread is created, passing in the clob, thread is started and then proceeds to try to read the stream, exceptions are seen
I'll attach the sample code.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
14 years, 9 months
[JBoss JIRA] Commented: (TEIID-1638) Add ability to query parent record info from child table query without joins for the Salesforce Translator
by Steven Hawkins (JIRA)
[ https://issues.jboss.org/browse/TEIID-1638?page=com.atlassian.jira.plugin... ]
Steven Hawkins commented on TEIID-1638:
---------------------------------------
The corresponding left outer join form should get pushed down as a relationship query.
e.g. something like: select c.id, c.name, c.accoundid, a.name from contact c left outer join account a on c.id = c.contactid
There are no plans at this time to support SOQL queries directly.
> Add ability to query parent record info from child table query without joins for the Salesforce Translator
> ----------------------------------------------------------------------------------------------------------
>
> Key: TEIID-1638
> URL: https://issues.jboss.org/browse/TEIID-1638
> Project: Teiid
> Issue Type: Feature Request
> Components: Salesforce Connector
> Affects Versions: 7.1.1
> Reporter: Debbie Steigner
> Labels: eds, salesforce
>
> Salesforce.com SOQL supports selecting parent record value through reference key name.
> We need to query parent record info from child table query. For example, if I have column parentid in Contact object which is reference key to account object, I can do in SOQL in salesforce.com something like this.
> Select c.Id, c.Name, c.AccountId, c.Account.Name
> from Contact c
> But if i want to do the same in EDS, I have to use Relationship Queries (join of two tables Account and Contact) which is huge impact on performance
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
14 years, 9 months