September 2018 - teiid-issues - Jboss List Archives

[JBoss JIRA] (TEIID-5471) string_agg should return null rather than empty over no values

by Steven Hawkins (JIRA)

[ https://issues.jboss.org/browse/TEIID-5471?page=com.atlassian.jira.plugin... ] Steven Hawkins resolved TEIID-5471. ----------------------------------- Resolution: Done Updated the output to be null. > string_agg should return null rather than empty over no values > -------------------------------------------------------------- > > Key: TEIID-5471 > URL: https://issues.jboss.org/browse/TEIID-5471 > Project: Teiid > Issue Type: Bug > Components: Query Engine > Reporter: Steven Hawkins > Assignee: Steven Hawkins > Fix For: 11.2 > > > For example the query: > select string_agg(col, ',') from (select cast(null as string) as col) x > Should return null rather than an empty string. -- This message was sent by Atlassian JIRA (v7.5.0#75005)

7 years, 6 months

1
0
0 / 0

[JBoss JIRA] (TEIID-5471) string_agg should return null rather than empty over no values

by Steven Hawkins (JIRA)

[ https://issues.jboss.org/browse/TEIID-5471?page=com.atlassian.jira.plugin... ] Steven Hawkins updated TEIID-5471: ---------------------------------- Fix Version/s: 11.1.1 11.0.3 > string_agg should return null rather than empty over no values > -------------------------------------------------------------- > > Key: TEIID-5471 > URL: https://issues.jboss.org/browse/TEIID-5471 > Project: Teiid > Issue Type: Bug > Components: Query Engine > Reporter: Steven Hawkins > Assignee: Steven Hawkins > Fix For: 11.2, 11.1.1, 11.0.3 > > > For example the query: > select string_agg(col, ',') from (select cast(null as string) as col) x > Should return null rather than an empty string. -- This message was sent by Atlassian JIRA (v7.5.0#75005)

7 years, 6 months

1
0
0 / 0

[JBoss JIRA] (TEIID-5469) Allow concat aggregates to work with over order by

by Steven Hawkins (JIRA)

[ https://issues.jboss.org/browse/TEIID-5469?page=com.atlassian.jira.plugin... ] Steven Hawkins resolved TEIID-5469. ----------------------------------- Resolution: Done With the new processing logic we support processing of concat aggs over an order by. > Allow concat aggregates to work with over order by > -------------------------------------------------- > > Key: TEIID-5469 > URL: https://issues.jboss.org/browse/TEIID-5469 > Project: Teiid > Issue Type: Enhancement > Components: Query Engine > Reporter: Steven Hawkins > Assignee: Steven Hawkins > Fix For: 11.2 > > > A validation was added to prevent the aggregates, such as string_agg, xmlagg, jsonarray_agg, etc. from working with an over order by clause - because most of them are destructive when the value is created and don't allow for rolling processing. -- This message was sent by Atlassian JIRA (v7.5.0#75005)

7 years, 6 months

1
0
0 / 0

[JBoss JIRA] (TEIID-5471) string_agg should return null rather than empty over no values

by Steven Hawkins (JIRA)

Steven Hawkins created TEIID-5471: ------------------------------------- Summary: string_agg should return null rather than empty over no values Key: TEIID-5471 URL: https://issues.jboss.org/browse/TEIID-5471 Project: Teiid Issue Type: Bug Components: Query Engine Reporter: Steven Hawkins Assignee: Steven Hawkins Fix For: 11.2 For example the query: select string_agg(col, ',') from (select cast(null as string) as col) x Should return null rather than an empty string. -- This message was sent by Atlassian JIRA (v7.5.0#75005)

7 years, 6 months

1
0
0 / 0

[JBoss JIRA] (TEIID-5467) Sorted tuple sources are not proactively cleaned in window function processing

by Steven Hawkins (JIRA)

[ https://issues.jboss.org/browse/TEIID-5467?page=com.atlassian.jira.plugin... ] Steven Hawkins resolved TEIID-5467. ----------------------------------- Fix Version/s: 11.1.1 Resolution: Done Added proactive cleanup for both the sort phase and processing phase in the event of failures. > Sorted tuple sources are not proactively cleaned in window function processing > ------------------------------------------------------------------------------ > > Key: TEIID-5467 > URL: https://issues.jboss.org/browse/TEIID-5467 > Project: Teiid > Issue Type: Quality Risk > Components: Query Engine > Reporter: Steven Hawkins > Assignee: Steven Hawkins > Fix For: 11.2, 11.1.1 > > > If window function processing completes normally the iteration over the sorted tuplebuffer is enough to remove it - since it's forward only. But an exception will cause the tuple buffer to wait for implicit cleanup. -- This message was sent by Atlassian JIRA (v7.5.0#75005)

7 years, 6 months

1
0
0 / 0

[JBoss JIRA] (TEIID-5470) Improve window function processing with order by over a literal

by Steven Hawkins (JIRA)

Steven Hawkins created TEIID-5470: ------------------------------------- Summary: Improve window function processing with order by over a literal Key: TEIID-5470 URL: https://issues.jboss.org/browse/TEIID-5470 Project: Teiid Issue Type: Enhancement Components: Query Engine Reporter: Steven Hawkins Assignee: Steven Hawkins Fix For: 11.2 We are sill processing the sort phase even when sorting over a literal. -- This message was sent by Atlassian JIRA (v7.5.0#75005)

7 years, 6 months

1
0
0 / 0

[JBoss JIRA] (TEIID-5469) Allow concat aggregates to work with over order by

by Steven Hawkins (JIRA)

Steven Hawkins created TEIID-5469: ------------------------------------- Summary: Allow concat aggregates to work with over order by Key: TEIID-5469 URL: https://issues.jboss.org/browse/TEIID-5469 Project: Teiid Issue Type: Enhancement Components: Query Engine Reporter: Steven Hawkins Assignee: Steven Hawkins Fix For: 11.2 A validation was added to prevent the aggregates, such as string_agg, xmlagg, jsonarray_agg, etc. from working with an over order by clause - because most of them are destructive when the value is created and don't allow for rolling processing. -- This message was sent by Atlassian JIRA (v7.5.0#75005)

7 years, 6 months

1
0
0 / 0

[JBoss JIRA] (TEIID-5394) Define how to utilize multiple vdbs

by Steven Hawkins (JIRA)

[ https://issues.jboss.org/browse/TEIID-5394?page=com.atlassian.jira.plugin... ] Steven Hawkins commented on TEIID-5394: --------------------------------------- > What other kinds of VDBs are you thinking about? I should have says works only with xml zips. It does not work with -vdb.xml, -vdb.ddl, and a zip with -vdb.ddl. > I do not think we have anyone just using the just DDL based vdb. Bram Gadeyne and several other users. > I do not think I follow this one, you mean because we are using the maven versioning scheme? Yes, just wanted to capture that requirement/limitation. > I did not realize we are doing this in the server code, will have to take peek. I mis-remembered about properties, they aren't merged in CompositeVDB, but we do also have visibility overrides to deal with. > Define how to utilize multiple vdbs > ----------------------------------- > > Key: TEIID-5394 > URL: https://issues.jboss.org/browse/TEIID-5394 > Project: Teiid > Issue Type: Sub-task > Components: Build/Kits > Reporter: Steven Hawkins > Assignee: Steven Hawkins > Fix For: 11.2 > > > Perhaps to support vdb imports, we should define how a build can incorporate multiple vdbs. -- This message was sent by Atlassian JIRA (v7.5.0#75005)

7 years, 6 months

1
0
0 / 0

[JBoss JIRA] (TEIID-5394) Define how to utilize multiple vdbs

by Ramesh Reddy (JIRA)

[ https://issues.jboss.org/browse/TEIID-5394?page=com.atlassian.jira.plugin... ] Ramesh Reddy commented on TEIID-5394: ------------------------------------- > only for xml based vdbs (ddl with the current restrictions is potentially supportable) What other kinds of VDBs are you thinking about? this should handle XML VDBs with external DDL files. I am expecting the original VDB in their project's "vdb" directory for this plugin, the contents will be in exploded format. i.e. if I were to take Designer's new VDB format with ddl files, I would place them in exploded format in my maven project. I do not think we have anyone just using the just DDL based vdb. > you have to create an appropriate uber config yaml for the importing vdb I am keeping all the configuration out of VDB layer. The previous version of the plugin by default copied the project-default.yaml file, but I commented out that for VDB building. Yes, you are correct, since we no longer have single server deployment model configuration needs to be (re)defined, I see that as an issue with this immutable deployment model than this VDB import process itself. > requires the use of semantic versioning I do not think I follow this one, you mean because we are using the maven versioning scheme? > It also may need to handle the merge of vdb properties. I did not realize we are doing this in the server code, will have to take peek. Should be a simple change. Same with data roles related changes looks like we need to do more introspection there than blind copy. > Having a child vdb with imports wouldn't be an issue correct? I have explicitly not allowed the nesting of imported VDBs. IMO, it is a matter of building multiple projects on the developer side, I believe that is exactly what are you saying too. > Define how to utilize multiple vdbs > ----------------------------------- > > Key: TEIID-5394 > URL: https://issues.jboss.org/browse/TEIID-5394 > Project: Teiid > Issue Type: Sub-task > Components: Build/Kits > Reporter: Steven Hawkins > Assignee: Steven Hawkins > Fix For: 11.2 > > > Perhaps to support vdb imports, we should define how a build can incorporate multiple vdbs. -- This message was sent by Atlassian JIRA (v7.5.0#75005)

7 years, 6 months

1
0
0 / 0

[JBoss JIRA] (TEIID-5394) Define how to utilize multiple vdbs

by Steven Hawkins (JIRA)

[ https://issues.jboss.org/browse/TEIID-5394?page=com.atlassian.jira.plugin... ] Steven Hawkins commented on TEIID-5394: --------------------------------------- The limitations will be: - only for xml based vdbs (ddl with the current restrictions is potentially supportable) - you have to create an appropriate uber config yaml for the importing vdb - requires the use of semantic versioning It also may need to handle the merge of vdb properties. It may also need to do more accounting with regards to what is an imported model. If you for example have a grant all in an imported vdb, the current server logic only lets that apply to the models from the imported vdb. Having a child vdb with imports wouldn't be an issue correct? Because the artifact built for that would itself be the merged version, which could then be remerged. > Define how to utilize multiple vdbs > ----------------------------------- > > Key: TEIID-5394 > URL: https://issues.jboss.org/browse/TEIID-5394 > Project: Teiid > Issue Type: Sub-task > Components: Build/Kits > Reporter: Steven Hawkins > Assignee: Steven Hawkins > Fix For: 11.2 > > > Perhaps to support vdb imports, we should define how a build can incorporate multiple vdbs. -- This message was sent by Atlassian JIRA (v7.5.0#75005)

7 years, 6 months

1
0
0 / 0

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

teiid-issues September 2018