March 2017 - teiid-issues - Jboss List Archives

[JBoss JIRA] (TEIID-4823) joining 2 procedures results in QueryPlannerException: TEIID30275 Join region with unsatisfied access patterns cannot be satisfied by the join criteria

by Steven Hawkins (JIRA)

[ https://issues.jboss.org/browse/TEIID-4823?page=com.atlassian.jira.plugin... ] Steven Hawkins commented on TEIID-4823: --------------------------------------- The root issue here is that the validation logic is not generally correct. If I suppress that exception, this scenario will plan/process just fine. > joining 2 procedures results in QueryPlannerException: TEIID30275 Join region with unsatisfied access patterns cannot be satisfied by the join criteria > -------------------------------------------------------------------------------------------------------------------------------------------------------- > > Key: TEIID-4823 > URL: https://issues.jboss.org/browse/TEIID-4823 > Project: Teiid > Issue Type: Bug > Affects Versions: 9.1.4 > Reporter: Bram Gadeyne > Assignee: Steven Hawkins > Attachments: plan1.txt, plan2.txt > > > I have a query like this where I join a regular table with a virtual procedure. > > {code:sql} > select gd.patientid, ip.production > from prod_P_GeneralData gd > join pr_patient_in_production ip on ip.patientid = gd.patientid > where gd.status = 1; > {code} > > This works fine and returns 58 rows with no null values. The pr_patient_in_production procedure has 1 input parameter called patientid (integer) and one output parameter called production (boolean) > > Adding a second procedure call results in an excepion: > > {code:sql} > select gd.patientid, lrs.datetime, lrs.varvalue as phtype > from prod_P_GeneralData gd > join pr_patient_in_production ip on ip.patientid = gd.patientid > join pr_retrieve_labres lrs on > lrs.patientid_in = gd.patientid > and lrs.production = ip.production > and lrs.resultgroup_in = 'pH-type' > and lrs.varvalue = 1 > where gd.status = 1 > order by lrs.datetime, lrs.entertime desc > {code} > > This pr_retrieve_labres procedure has patientid_in (integer), production (boolean) and resultgroup_in(string) as input parameters and also some output parameters. > > The exception is: Error: TEIID30275 Remote org.teiid.api.exception.query.QueryPlannerException: TEIID30275 Join region with unsatisfied access patterns cannot be satisfied by the join criteria, Access patterns: [[Access Pattern: Unsatisfied [ip.patientid] History [[ip.patientid]]], [Access Pattern: Unsatisfied [lrs.production, lrs.patientid_in] History [[lrs.patientid_in, lrs.production, lrs.resultgroup_in]]]] -- This message was sent by Atlassian JIRA (v7.2.3#72005)

7 years, 9 months

1
0
0 / 0

[JBoss JIRA] (TEIID-4827) Update the readme to indicate that Java 1.8 is required for building

by Steven Hawkins (JIRA)

[ https://issues.jboss.org/browse/TEIID-4827?page=com.atlassian.jira.plugin... ] Steven Hawkins commented on TEIID-4827: --------------------------------------- Ok, so now the readme has been updated and the build requires JDK 1.8. Untill we change the parent pom / wildfly version expectation we'll leave the source and target as 1.7. > Update the readme to indicate that Java 1.8 is required for building > -------------------------------------------------------------------- > > Key: TEIID-4827 > URL: https://issues.jboss.org/browse/TEIID-4827 > Project: Teiid > Issue Type: Bug > Reporter: Steven Hawkins > Assignee: Steven Hawkins > Fix For: 9.3 > > > There are several places where we are implicitly depending upon 1.8 features to build - such as the larger key size from TestDhKeyGenerator. These issues do not affect compatibility with a 1.7 runtime. -- This message was sent by Atlassian JIRA (v7.2.3#72005)

7 years, 9 months

1
0
0 / 0

[JBoss JIRA] (TEIID-4827) Update the readme to indicate that Java 1.8 is required for building

by Ramesh Reddy (JIRA)

[ https://issues.jboss.org/browse/TEIID-4827?page=com.atlassian.jira.plugin... ] Ramesh Reddy commented on TEIID-4827: ------------------------------------- I think we should, even the Infinispan translator I am working on needs a 1.8 due to Infinispan dependencies. > Update the readme to indicate that Java 1.8 is required for building > -------------------------------------------------------------------- > > Key: TEIID-4827 > URL: https://issues.jboss.org/browse/TEIID-4827 > Project: Teiid > Issue Type: Bug > Reporter: Steven Hawkins > Assignee: Steven Hawkins > Fix For: 9.3 > > > There are several places where we are implicitly depending upon 1.8 features to build - such as the larger key size from TestDhKeyGenerator. These issues do not affect compatibility with a 1.7 runtime. -- This message was sent by Atlassian JIRA (v7.2.3#72005)

7 years, 9 months

1
0
0 / 0

[JBoss JIRA] (TEIID-4825) Google translator throws Column doesn't exist in the worksheet

by Steven Hawkins (JIRA)

[ https://issues.jboss.org/browse/TEIID-4825?page=com.atlassian.jira.plugin... ] Steven Hawkins edited comment on TEIID-4825 at 3/24/17 3:23 PM: ---------------------------------------------------------------- This is similar to TEIID-4672. Essentially the google-api has it's own rules as to when a column value is header or not. If there is just a header row or even just string data - it's just treated as a data row. With numeric content - it's automatically a header. This is a little thorny for us to undo. Even if we make the metadata match what we expect, queries that expect a header, such as update, operations will fail. I'll look again to see if there is a quick resolution. Otherwise it's simply the case that to google there is no header for the sheet in question and our metadata shouldn't treat it as having one. was (Author: shawkins): This is similar to TEIID-4672. Essentially the google-api has it's own rules as to when a column value is header or not. If there is just a header row or even just string data - it's just treated as a data row. With numeric content - it's automatically a header. This is a little thorny for us to undo. As even if we make the metadata match what we expect queries that expect a header, such as update operations will fail. I'll look again to see if there is a quick resolution. Otherwise it's simply the case that to google there is no header for the sheet in question and our metadata shouldn't treat it as having one. > Google translator throws Column doesn't exist in the worksheet > -------------------------------------------------------------- > > Key: TEIID-4825 > URL: https://issues.jboss.org/browse/TEIID-4825 > Project: Teiid > Issue Type: Bug > Components: Misc. Connectors > Affects Versions: 8.12.10.6_3 > Reporter: Lucie Fabrikova > Assignee: Steven Hawkins > Fix For: 9.3 > > Attachments: gs-export-vdb.xml, server.log-dv635cr1 > > > If sheet (WikiData9) contains only string type columns, queries to that sheet fail with 'org.teiid.resource.adapter.google.common.SpreadsheetOperationException: Column link doesn't exist in the worksheet WikiData9' > If I added e.g. numeric or boolean column, it worked ok... -- This message was sent by Atlassian JIRA (v7.2.3#72005)

7 years, 9 months

1
0
0 / 0

[JBoss JIRA] (TEIID-4821) Debug Plan is missing Annotation about JOIN Pushdown

by Steven Hawkins (JIRA)

[ https://issues.jboss.org/browse/TEIID-4821?page=com.atlassian.jira.plugin... ] Steven Hawkins resolved TEIID-4821. ----------------------------------- Fix Version/s: 9.3 Resolution: Done Added an annotation specifically for the access pattern case. > Debug Plan is missing Annotation about JOIN Pushdown > ----------------------------------------------------- > > Key: TEIID-4821 > URL: https://issues.jboss.org/browse/TEIID-4821 > Project: Teiid > Issue Type: Quality Risk > Components: Query Engine > Reporter: Ramesh Reddy > Assignee: Steven Hawkins > Fix For: 9.3 > > > When a translator supports the JOIN pushdown and the submitted query is not using the JOIN pushdown, there needs to be description in the plan as to why query planner did not include the JOIN in the pushdown query to translator. Currently this is not available. -- This message was sent by Atlassian JIRA (v7.2.3#72005)

7 years, 9 months

1
0
0 / 0

[JBoss JIRA] (TEIID-4829) odata4 fails to return data on top of mongoDB view model

by Michael Echevarria (JIRA)

Michael Echevarria created TEIID-4829: ----------------------------------------- Summary: odata4 fails to return data on top of mongoDB view model Key: TEIID-4829 URL: https://issues.jboss.org/browse/TEIID-4829 Project: Teiid Issue Type: Bug Components: Misc. Connectors Affects Versions: 8.12.9.6_3 Environment: Noticed in JDV 6.3.4 Reporter: Michael Echevarria Assignee: Steven Hawkins Priority: Critical Attachments: mongo-view-error.png, odata2-works.png, odata4-error.png Description of problem: Accessing either an exclusive view model or combined view model for a mongoDB datasource via odata4 results in an error. The data returns via odata2, jdbc and preview in the designer Version-Release number of selected component (if applicable): JDV 6.3.4 Actual results: Property mapping error Expected results: Model data -- This message was sent by Atlassian JIRA (v7.2.3#72005)

7 years, 9 months

1
0
0 / 0

[JBoss JIRA] (TEIID-4821) Debug Plan is missing Annotation about JOIN Pushdown

by Steven Hawkins (JIRA)

[ https://issues.jboss.org/browse/TEIID-4821?page=com.atlassian.jira.plugin... ] Steven Hawkins updated TEIID-4821: ---------------------------------- Issue Type: Quality Risk (was: Enhancement) > Debug Plan is missing Annotation about JOIN Pushdown > ----------------------------------------------------- > > Key: TEIID-4821 > URL: https://issues.jboss.org/browse/TEIID-4821 > Project: Teiid > Issue Type: Quality Risk > Components: Query Engine > Reporter: Ramesh Reddy > Assignee: Steven Hawkins > > When a translator supports the JOIN pushdown and the submitted query is not using the JOIN pushdown, there needs to be description in the plan as to why query planner did not include the JOIN in the pushdown query to translator. Currently this is not available. -- This message was sent by Atlassian JIRA (v7.2.3#72005)

7 years, 9 months

1
0
0 / 0

[JBoss JIRA] (TEIID-4825) Google translator throws Column doesn't exist in the worksheet

by Steven Hawkins (JIRA)

[ https://issues.jboss.org/browse/TEIID-4825?page=com.atlassian.jira.plugin... ] Steven Hawkins commented on TEIID-4825: --------------------------------------- This is similar to TEIID-4672. Essentially the google-api has it's own rules as to when a column value is header or not. If there is just a header row or even just string data - it's just treated as a data row. With numeric content - it's automatically a header. This is a little thorny for us to undo. As even if we make the metadata match what we expect queries that expect a header, such as update operations will fail. I'll look again to see if there is a quick resolution. Otherwise it's simply the case that to google there is no header for the sheet in question and our metadata shouldn't treat it as having one. > Google translator throws Column doesn't exist in the worksheet > -------------------------------------------------------------- > > Key: TEIID-4825 > URL: https://issues.jboss.org/browse/TEIID-4825 > Project: Teiid > Issue Type: Bug > Components: Misc. Connectors > Affects Versions: 8.12.10.6_3 > Reporter: Lucie Fabrikova > Assignee: Steven Hawkins > Fix For: 9.3 > > Attachments: gs-export-vdb.xml, server.log-dv635cr1 > > > If sheet (WikiData9) contains only string type columns, queries to that sheet fail with 'org.teiid.resource.adapter.google.common.SpreadsheetOperationException: Column link doesn't exist in the worksheet WikiData9' > If I added e.g. numeric or boolean column, it worked ok... -- This message was sent by Atlassian JIRA (v7.2.3#72005)

7 years, 9 months

1
0
0 / 0

[JBoss JIRA] (TEIID-4824) Add unique id as a configuration property in the Google Resource adapter

by Steven Hawkins (JIRA)

[ https://issues.jboss.org/browse/TEIID-4824?page=com.atlassian.jira.plugin... ] Steven Hawkins updated TEIID-4824: ---------------------------------- Component/s: Misc. Connectors Fix Version/s: 9.3 This looks straight-forward, but we'll need a new resource adapter property. It can either be a boolean flag to indicate a key is being used as the name, or a separate string field that is the key. > Add unique id as a configuration property in the Google Resource adapter > ------------------------------------------------------------------------ > > Key: TEIID-4824 > URL: https://issues.jboss.org/browse/TEIID-4824 > Project: Teiid > Issue Type: Enhancement > Components: Misc. Connectors > Affects Versions: 8.12.8.6_3 > Reporter: Debbie Steigner > Assignee: Steven Hawkins > Fix For: 9.3 > > > Currently, only the google spreadsheet name can be used in the Google Resource adapter instead of the unique id of the document. If there is more than one document with same name in multiple folders the wrong file may be picked up by the resource adapter. Is it possible to use unique document ID instead of spreadsheet name for google data source -- This message was sent by Atlassian JIRA (v7.2.3#72005)

7 years, 9 months

1
0
0 / 0

[JBoss JIRA] (TEIID-2465) Add additional size limitations

by Steven Hawkins (JIRA)

[ https://issues.jboss.org/browse/TEIID-2465?page=com.atlassian.jira.plugin... ] Steven Hawkins resolved TEIID-2465. ----------------------------------- Resolution: Done Tweaked the maximum for a tuplebuffer and added documentation in the memory management section about the limits. > Add additional size limitations > ------------------------------- > > Key: TEIID-2465 > URL: https://issues.jboss.org/browse/TEIID-2465 > Project: Teiid > Issue Type: Quality Risk > Components: Query Engine > Reporter: Steven Hawkins > Assignee: Steven Hawkins > Fix For: 9.3 > > > We should introduce a limit to lob (which is currently only correct for up to 2 gigs) and single tuplebuffer results in storage as to prevent too much resource consumption from a single use. -- This message was sent by Atlassian JIRA (v7.2.3#72005)

7 years, 9 months

1
0
0 / 0

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

teiid-issues March 2017