September 2018 - teiid-issues - Jboss List Archives

[JBoss JIRA] (TEIID-5473) Make JDBCMetadataProcessor more extensible with protected methods

by Ken Geis (JIRA)

[ https://issues.jboss.org/browse/TEIID-5473?page=com.atlassian.jira.plugin... ] Ken Geis commented on TEIID-5473: --------------------------------- I'm using the Oracle driver. I just sampled (n=1) and found that without {{-DuseFetchSizeWithLongColumn=true}}, it takes 2m48s. With it, it takes 2m11s. I'm not comfortable setting this property application-wide, but I think it could be [set on the individual connection|https://docs.oracle.com/javase/8/docs/api/java/sql/Connection....] if I override {{process(..)}} or {{getConnectorMetadata(..)}}. Yes, I can now see a path for trying a different Oracle-specific method for fetching index information by returning false in {{getIndexInfoForTable}} and then using my own logic in the overridden {{getConnectorMetadata(..)}}. If you have any hesitation in modifying the API per my request, let me test out my hypothesis first. > Make JDBCMetadataProcessor more extensible with protected methods > ----------------------------------------------------------------- > > Key: TEIID-5473 > URL: https://issues.jboss.org/browse/TEIID-5473 > Project: Teiid > Issue Type: Feature Request > Components: JDBC Connector > Affects Versions: 11.1 > Reporter: Ken Geis > Assignee: Steven Hawkins > Fix For: 11.2 > > > I believe I can write a JDBCMetadataProcessor that works faster against my Oracle database. The important part to override is in getIndexes(..). This method is package-private, so I cannot override it. It would be useful to me if the method (and maybe others) were made protected so that they can be overridden. -- This message was sent by Atlassian JIRA (v7.5.0#75005)

7 years, 1 month

1
0
0 / 0

[JBoss JIRA] (TEIID-5400) Further normalize source queries

by Steven Hawkins (JIRA)

[ https://issues.jboss.org/browse/TEIID-5400?page=com.atlassian.jira.plugin... ] Work on TEIID-5400 started by Steven Hawkins. --------------------------------------------- > Further normalize source queries > -------------------------------- > > Key: TEIID-5400 > URL: https://issues.jboss.org/browse/TEIID-5400 > Project: Teiid > Issue Type: Sub-task > Components: Query Engine > Reporter: Steven Hawkins > Assignee: Steven Hawkins > Fix For: 11.2 > > > To make better use of cache entries source queries should be more normalized, such as alpha sorting the projected columns. -- This message was sent by Atlassian JIRA (v7.5.0#75005)

7 years, 1 month

1
0
0 / 0

[JBoss JIRA] (TEIID-5473) Make JDBCMetadataProcessor more extensible with protected methods

by Steven Hawkins (JIRA)

[ https://issues.jboss.org/browse/TEIID-5473?page=com.atlassian.jira.plugin... ] Steven Hawkins commented on TEIID-5473: --------------------------------------- There is a getIndexInfoForTable protected method that needs a better java doc that should suffice. If you override it you just need to return false to suppress the default logic from running. If overriding that method can suffice, then I'll add an analogous method for foreign keys as well. Ivan was using the progress driver for which there is the kb article: https://knowledgebase.progress.com/articles/Article/Calls-using-getImport... For the regular driver we've seen that useFetchSizeWithLongColumn=true https://docs.oracle.com/cd/E11882_01/appdev.112/e13995/oracle/jdbc/Oracle... can improve things as well. > Make JDBCMetadataProcessor more extensible with protected methods > ----------------------------------------------------------------- > > Key: TEIID-5473 > URL: https://issues.jboss.org/browse/TEIID-5473 > Project: Teiid > Issue Type: Feature Request > Components: JDBC Connector > Affects Versions: 11.1 > Reporter: Ken Geis > Assignee: Steven Hawkins > Fix For: 11.2 > > > I believe I can write a JDBCMetadataProcessor that works faster against my Oracle database. The important part to override is in getIndexes(..). This method is package-private, so I cannot override it. It would be useful to me if the method (and maybe others) were made protected so that they can be overridden. -- This message was sent by Atlassian JIRA (v7.5.0#75005)

7 years, 1 month

1
0
0 / 0

[JBoss JIRA] (TEIID-5473) Make JDBCMetadataProcessor more extensible with protected methods

by Steven Hawkins (JIRA)

[ https://issues.jboss.org/browse/TEIID-5473?page=com.atlassian.jira.plugin... ] Work on TEIID-5473 started by Steven Hawkins. --------------------------------------------- > Make JDBCMetadataProcessor more extensible with protected methods > ----------------------------------------------------------------- > > Key: TEIID-5473 > URL: https://issues.jboss.org/browse/TEIID-5473 > Project: Teiid > Issue Type: Feature Request > Components: JDBC Connector > Affects Versions: 11.1 > Reporter: Ken Geis > Assignee: Steven Hawkins > Fix For: 11.2 > > > I believe I can write a JDBCMetadataProcessor that works faster against my Oracle database. The important part to override is in getIndexes(..). This method is package-private, so I cannot override it. It would be useful to me if the method (and maybe others) were made protected so that they can be overridden. -- This message was sent by Atlassian JIRA (v7.5.0#75005)

7 years, 1 month

1
0
0 / 0

[JBoss JIRA] (TEIID-3931) Add support for odata service geospatial handling

by Steven Hawkins (JIRA)

[ https://issues.jboss.org/browse/TEIID-3931?page=com.atlassian.jira.plugin... ] Steven Hawkins commented on TEIID-3931: --------------------------------------- Given the added importance to odata I'm going to take a fresh look at this to see if we can have more robust support. > Add support for odata service geospatial handling > ------------------------------------------------- > > Key: TEIID-3931 > URL: https://issues.jboss.org/browse/TEIID-3931 > Project: Teiid > Issue Type: Enhancement > Components: OData > Reporter: Steven Hawkins > Assignee: Steven Hawkins > Fix For: 11.2 > > > We should use the Edm.Geometry type for the Teiid geometry type (we can also use system metadata to see if it's appropriate to use a geometry subtype) along with any geospatial operations that odata and Teiid support. -- This message was sent by Atlassian JIRA (v7.5.0#75005)

7 years, 1 month

1
0
0 / 0

[JBoss JIRA] (TEIID-3931) Add support for odata service geospatial handling

by Steven Hawkins (JIRA)

[ https://issues.jboss.org/browse/TEIID-3931?page=com.atlassian.jira.plugin... ] Work on TEIID-3931 started by Steven Hawkins. --------------------------------------------- > Add support for odata service geospatial handling > ------------------------------------------------- > > Key: TEIID-3931 > URL: https://issues.jboss.org/browse/TEIID-3931 > Project: Teiid > Issue Type: Enhancement > Components: OData > Reporter: Steven Hawkins > Assignee: Steven Hawkins > Fix For: 11.2 > > > We should use the Edm.Geometry type for the Teiid geometry type (we can also use system metadata to see if it's appropriate to use a geometry subtype) along with any geospatial operations that odata and Teiid support. -- This message was sent by Atlassian JIRA (v7.5.0#75005)

7 years, 1 month

1
0
0 / 0

[JBoss JIRA] (TEIID-5478) Add engine support for nth_value

by Steven Hawkins (JIRA)

Steven Hawkins created TEIID-5478: ------------------------------------- Summary: Add engine support for nth_value Key: TEIID-5478 URL: https://issues.jboss.org/browse/TEIID-5478 Project: Teiid Issue Type: Sub-task Components: Query Engine Reporter: Steven Hawkins Assignee: Steven Hawkins Fix For: 11.2 Due to TEIID-5475 engine support for nth_value was removed. It should be added back with frame clause support. -- This message was sent by Atlassian JIRA (v7.5.0#75005)

7 years, 1 month

1
0
0 / 0

[JBoss JIRA] (TEIID-5475) nth_value is incorrect for some initial values

by Steven Hawkins (JIRA)

[ https://issues.jboss.org/browse/TEIID-5475?page=com.atlassian.jira.plugin... ] Steven Hawkins resolved TEIID-5475. ----------------------------------- Resolution: Done Resolved by removing engine support for now. Will add corrected support later. > nth_value is incorrect for some initial values > ---------------------------------------------- > > Key: TEIID-5475 > URL: https://issues.jboss.org/browse/TEIID-5475 > Project: Teiid > Issue Type: Bug > Components: Query Engine > Reporter: Steven Hawkins > Assignee: Steven Hawkins > Priority: Blocker > Fix For: 11.2, 11.1.1 > > > Nth_value is returning null when not expected. For example: > select e1, nth_value(e1, 2) over (order by e1 nulls last) from g1 > with our default data set should return 'a' for each nth_value, but instead returns null for one of the 'a' rows. The issue is that we should be considering the indexing to be relative from the start of the peer group in the frame rather than from the current index of the row. -- This message was sent by Atlassian JIRA (v7.5.0#75005)

7 years, 1 month

1
0
0 / 0

[JBoss JIRA] (TEIID-5473) Make JDBCMetadataProcessor more extensible with protected methods

by Steven Hawkins (JIRA)

[ https://issues.jboss.org/browse/TEIID-5473?page=com.atlassian.jira.plugin... ] Steven Hawkins updated TEIID-5473: ---------------------------------- Fix Version/s: 11.2 > Make JDBCMetadataProcessor more extensible with protected methods > ----------------------------------------------------------------- > > Key: TEIID-5473 > URL: https://issues.jboss.org/browse/TEIID-5473 > Project: Teiid > Issue Type: Feature Request > Components: JDBC Connector > Affects Versions: 11.1 > Reporter: Ken Geis > Assignee: Steven Hawkins > Fix For: 11.2 > > > I believe I can write a JDBCMetadataProcessor that works faster against my Oracle database. The important part to override is in getIndexes(..). This method is package-private, so I cannot override it. It would be useful to me if the method (and maybe others) were made protected so that they can be overridden. -- This message was sent by Atlassian JIRA (v7.5.0#75005)

7 years, 1 month

1
0
0 / 0

[JBoss JIRA] (TEIID-5475) nth_value is incorrect for some initial values

by Steven Hawkins (JIRA)

[ https://issues.jboss.org/browse/TEIID-5475?page=com.atlassian.jira.plugin... ] Work on TEIID-5475 started by Steven Hawkins. --------------------------------------------- > nth_value is incorrect for some initial values > ---------------------------------------------- > > Key: TEIID-5475 > URL: https://issues.jboss.org/browse/TEIID-5475 > Project: Teiid > Issue Type: Bug > Components: Query Engine > Reporter: Steven Hawkins > Assignee: Steven Hawkins > Priority: Blocker > Fix For: 11.2, 11.1.1 > > > Nth_value is returning null when not expected. For example: > select e1, nth_value(e1, 2) over (order by e1 nulls last) from g1 > with our default data set should return 'a' for each nth_value, but instead returns null for one of the 'a' rows. The issue is that we should be considering the indexing to be relative from the start of the peer group in the frame rather than from the current index of the row. -- This message was sent by Atlassian JIRA (v7.5.0#75005)

7 years, 1 month

1
0
0 / 0

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

teiid-issues September 2018