[JBoss JIRA] Created: (TEIID-914) OnWrapTransactionTests.testMultipleSourceTimeout test fails due to due to execpted error message is different
by Van Halbert (JIRA)
OnWrapTransactionTests.testMultipleSourceTimeout test fails due to due to execpted error message is different
-------------------------------------------------------------------------------------------------------------
Key: TEIID-914
URL: https://jira.jboss.org/jira/browse/TEIID-914
Project: Teiid
Issue Type: Bug
Components: Integration Tests
Affects Versions: 7.0
Reporter: Van Halbert
Assignee: Steven Hawkins
Priority: Minor
The OnWrapTransactionTests.testMultipleSourceTimeout test fails due to expected message is different. It appears the upstream error being thrown by sendRequestMessageAndWait(..) in MMStatement.executeSql(String[] commands, boolean isBatchedCommand, Boolean requiresResultSet) has changed. If I take the message from the caught exception and prepend to the created message and use that message in the MMSQLException throw, then the test passes.
Also, another problem with the current error message is that it gives no indication as to the problem why the statement cannot run, it only provides a general message, ilke the following:
"Error trying to execute a statement [insert into pm1.g1 (e1, e2) values(600,'600'), insert into pm1.g2 (e1, e2) values (600,'600') .........."
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: https://jira.jboss.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira
15 years, 8 months
[JBoss JIRA] Closed: (TEIID-968) Timestamp literals with fractional seconds result in source query execution performance when being compared to less percise time types
by Steve Hawkins (JIRA)
[ https://jira.jboss.org/browse/TEIID-968?page=com.atlassian.jira.plugin.sy... ]
Steve Hawkins closed TEIID-968.
-------------------------------
> Timestamp literals with fractional seconds result in source query execution performance when being compared to less percise time types
> --------------------------------------------------------------------------------------------------------------------------------------
>
> Key: TEIID-968
> URL: https://jira.jboss.org/browse/TEIID-968
> Project: Teiid
> Issue Type: Feature Request
> Affects Versions: 7.1
> Environment: MM 5.5.3
> Oracle 10i
> Oracle ANSI Connector
> Reporter: Larry O'Leary
> Assignee: Steve Hawkins
> Fix For: 7.1
>
> Attachments: mylyn-context.zip
>
> Original Estimate: 0 minutes
> Remaining Estimate: 0 minutes
>
> Seeing that the only "time" type for MetaMatrix is timestamp and Oracle uses DATE to store time and TIMESTAMP to store time with fractional seconds we need a method of sending a timestamp as a literal that Oracle can use as a DATE instead of requiring the comparison to be normalized to TIMESTAMP as this results in a huge performance hit within Oracle.
> It is possible to change this behaviour by actually setting a date only column to MetaMatrix date in the model but this does not account for situations where a DATE column in Oracle holds date and time information. In those cases, I would still need to use timestamp for MetaMatrix and this would result in a huge performance impact.
> Oracle will automatically use the widest type possible when doing comparisons so if I store date only data in a DATE column and then attempt to query using criteria on that column such ad {d'2000-01-01'} the date literal gets converted to {ts'2000-01-01 00:00:00.0'} prior to it being sent to Oracle. This will result in Oracle converting my DATE column into a TIMESTAMP column to handle the comparison.
> The current DataDirect driver we are using actually accounts for this issue when dealing with timestamp literals. Basically, if the timestamp literal does not include fractional seconds, the driver will pass it to Oracle as a date instead of a timestamp. This greatly improves performance in these situations.
> So, if the current logic is kept, but we simply truncate the timestamp literal to remove the fractional seconds, we get the advantage of the driver performing better type mapping in this situation.
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: https://jira.jboss.org/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira
15 years, 8 months