[JBoss JIRA] (TEIID-5311) Sybase implicit Convertion of String value 'yyyy-MM-dd HH:mm' into a Date value Causes mismatches in JDV Date comparison
by Steven Hawkins (JIRA)
[ https://issues.jboss.org/browse/TEIID-5311?page=com.atlassian.jira.plugin... ]
Work on TEIID-5311 stopped by Steven Hawkins.
---------------------------------------------
> Sybase implicit Convertion of String value 'yyyy-MM-dd HH:mm' into a Date value Causes mismatches in JDV Date comparison
> ------------------------------------------------------------------------------------------------------------------------
>
> Key: TEIID-5311
> URL: https://issues.jboss.org/browse/TEIID-5311
> Project: Teiid
> Issue Type: Enhancement
> Components: Query Engine
> Reporter: Anu Saji
> Assignee: Steven Hawkins
> Priority: Minor
> Fix For: 10.x
>
> Attachments: eniqpreparser_v4.zip
>
>
> Sybase implicitly converts a String value 'yyyy-MM-dd HH:mm' into a Date value.
> This format is not standard and this conversion is specific to Sybase.
> As JDBC/Java standard, JDV implicitly converts
> 'yyyy-MM-dd HH:mm:ss' into a Timestamp value
> 'yyyy-MM-dd' into a Date value
> So JDV does not considered 'yyyy-MM-dd HH:mm' as a Date value and does not proceed to any implicit conversion. 'yyyy-MM-dd HH:mm' remains like a String value which partially matches with Date comparison
> Constraint
> We cannot modify the incoming query
--
This message was sent by Atlassian JIRA
(v7.5.0#75005)
6 years, 8 months
[JBoss JIRA] (TEIID-5311) Sybase implicit Convertion of String value 'yyyy-MM-dd HH:mm' into a Date value Causes mismatches in JDV Date comparison
by Steven Hawkins (JIRA)
[ https://issues.jboss.org/browse/TEIID-5311?page=com.atlassian.jira.plugin... ]
Steven Hawkins updated TEIID-5311:
----------------------------------
Fix Version/s: 10.x
(was: 10.3)
> Sybase implicit Convertion of String value 'yyyy-MM-dd HH:mm' into a Date value Causes mismatches in JDV Date comparison
> ------------------------------------------------------------------------------------------------------------------------
>
> Key: TEIID-5311
> URL: https://issues.jboss.org/browse/TEIID-5311
> Project: Teiid
> Issue Type: Enhancement
> Components: Query Engine
> Reporter: Anu Saji
> Assignee: Steven Hawkins
> Priority: Minor
> Fix For: 10.x
>
> Attachments: eniqpreparser_v4.zip
>
>
> Sybase implicitly converts a String value 'yyyy-MM-dd HH:mm' into a Date value.
> This format is not standard and this conversion is specific to Sybase.
> As JDBC/Java standard, JDV implicitly converts
> 'yyyy-MM-dd HH:mm:ss' into a Timestamp value
> 'yyyy-MM-dd' into a Date value
> So JDV does not considered 'yyyy-MM-dd HH:mm' as a Date value and does not proceed to any implicit conversion. 'yyyy-MM-dd HH:mm' remains like a String value which partially matches with Date comparison
> Constraint
> We cannot modify the incoming query
--
This message was sent by Atlassian JIRA
(v7.5.0#75005)
6 years, 8 months
[JBoss JIRA] (TEIID-4760) optional object name manipulations while executing IMPORT FOREIGN SCHEMA
by Steven Hawkins (JIRA)
[ https://issues.jboss.org/browse/TEIID-4760?page=com.atlassian.jira.plugin... ]
Steven Hawkins resolved TEIID-4760.
-----------------------------------
Resolution: Done
Added importer.nameFormat that accepts a Java format string so that it can easily handle prefix and suffix operations.
> optional object name manipulations while executing IMPORT FOREIGN SCHEMA
> ------------------------------------------------------------------------
>
> Key: TEIID-4760
> URL: https://issues.jboss.org/browse/TEIID-4760
> Project: Teiid
> Issue Type: Feature Request
> Components: Misc. Connectors
> Reporter: Bram Gadeyne
> Assignee: Steven Hawkins
> Labels: ddl, import, import_wizard
> Fix For: 10.3
>
>
> While executing IMPORT FOREIGN SCHEMA, table (and other object) names are added to the schema with their original names.
> In our previous virtual database we imported two schema's that contain the same objects but one is a production and another is a historical database.
> We then change the name for e.g. table1 to prod_table1 or wh_table1 according to the schema it came from. Some software that interacts with teiid can not distinguish between different schemas and can not handle duplicate table names. So for this reason, this would also come in handy.
> It would come in handy if it is possible to add a prefix to names while importing objects.
--
This message was sent by Atlassian JIRA
(v7.5.0#75005)
6 years, 8 months
[JBoss JIRA] (TEIID-5096) Using /*+ MAKEDEP */ blocks the deploy proces when using DDL based vdb
by Bram Gadeyne (JIRA)
[ https://issues.jboss.org/browse/TEIID-5096?page=com.atlassian.jira.plugin... ]
Bram Gadeyne commented on TEIID-5096:
-------------------------------------
PM,
I'm now in 10.3 and I'm currently using some /*+ MAKEDEP */ hints in my ddl vdb. The issue hasn't shown up on this version.
I'm using the jboss cli to deploy the vdb and there are certainly no simultanious deployments.
Now you mention it, I was experiencing some network issues from time to time but I can't confirm that they were happening while this issue occured.
I'm afraid I will not be able to provide you with more information at this point in time. If it happens again I'll try and debug this more profound.
> Using /*+ MAKEDEP */ blocks the deploy proces when using DDL based vdb
> ----------------------------------------------------------------------
>
> Key: TEIID-5096
> URL: https://issues.jboss.org/browse/TEIID-5096
> Project: Teiid
> Issue Type: Bug
> Affects Versions: 9.3.3
> Reporter: Bram Gadeyne
> Assignee: Steven Hawkins
> Attachments: threaddump.txt
>
>
> Hi,
> I've created a vdb that is depoyed using a vdb-ddl.ddl file.
> The deployment just stops and returned no error message when adding the SQL part below. Subsequent deployments are not handles. It seems like the deployment process just hangs. The only way to resolve this was to stop the wildfly server, remove the latest added content from the standalone/data/content directory and restart wildfly.
> {code:sql}
> CREATE VIEW tv_retrieve_monvals(
> admissionid integer not null,
> VariableID integer not null,
> Datetime timestamp not null,
> Entertime timestamp not null,
> varvalue double not null,
> primary key (admissionid, VariableID, Datetime)
> )
> AS
> SELECT v.PatientID AS admissionid, v.VariableID, v.Datetime, v.Entertime, v."Value" AS varvalue
> FROM (
> SELECT mv.PatientID, mv.VariableID, mv.Datetime, mv.Entertime, mv."Value",
> ROW_NUMBER() OVER (PARTITION BY mv.PatientID, mv.VariableID, mv.Datetime ORDER BY mv.Entertime DESC) AS rang
> FROM izisprod.P_GeneralData AS gd
> INNER JOIN /*+ MAKEDEP */ izisprod.P_MonVals AS mv ON
> gd.PatientID = mv.PatientID AND
> bitand(mv.Status, 8) = 8 AND
> bitand(mv.Status, 2) <> 2
> WHERE gd.Status = 1 OR (gd.Status >= 4 AND gd.Status <> 5)
> UNION
> SELECT mv.PatientID, mv.VariableID, mv.Datetime, mv.Entertime, mv."Value",
> ROW_NUMBER() OVER (PARTITION BY mv.PatientID, mv.VariableID, mv.Datetime ORDER BY mv.Entertime DESC) AS rang
> FROM iziswh.P_GeneralData AS gd
> INNER JOIN /*+ MAKEDEP */ iziswh.P_MonVals AS mv ON
> gd.PatientID = mv.PatientID AND
> bitand(mv.Status, 8) = 8 AND
> bitand(mv.Status, 2) <> 2
> ) AS v
> WHERE v.rang = 1;
> {code}
> Removing the /*+ MAKDEP */ references resolves the issue.
> I've tried creating a small ddl script that uses /*+ MAKEDEP */ to reproduce this issue but this script does deploy so currently I can not make a small example.
--
This message was sent by Atlassian JIRA
(v7.5.0#75005)
6 years, 8 months