[ https://issues.apache.org/jira/browse/IGNITE-7785?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16378534#comment-16378534 ]
Alexander Paschenko commented on IGNITE-7785: --------------------------------------------- [~pvinokurov], Initial query is a bit weird as long as it has union of a table with itself. In absence of {{UNION ALL}} clause, that subquery {{(select id from company union all select id from company)}} should give same result as single {{SELECT id FROM Company}}, and thus in initial case that UNION is redundant - if you remove UNION, query works. However, if you add ALL to UNION, query works - however, in this case UNION has duplicates just like it should. I will also have a look at second query. > SQL query with COUNT and UNION in sub-query produces JdbcSQLException > --------------------------------------------------------------------- > > Key: IGNITE-7785 > URL: https://issues.apache.org/jira/browse/IGNITE-7785 > Project: Ignite > Issue Type: Bug > Components: sql > Affects Versions: 2.1, 2.3 > Reporter: Pavel Vinokurov > Priority: Major > > SQL initial script: > CREATE TABLE Person(id INTEGER PRIMARY KEY, company_id INTEGER); > CREATE TABLE Company(id INTEGER PRIMARY KEY, name VARCHAR); > INSERT INTO Person(id,company_id) VALUES (1, 1), (2, 2), (3, 3); > INSERT INTO Company(id,name) VALUES (1,'n1'), (2,'n2'), (3,'n3'); > SQL Query: > SELECT count(1) FROM person p > LEFT JOIN (select id from company union select id from company) as c on > c.id=p.company_id > JDBC Exception: > Caused by: org.h2.jdbc.JdbcSQLException: Column "P__Z0.COMPANY_ID" must be in > the GROUP BY list; SQL statement: > SELECT > P__Z0.COMPANY_ID __C0_0, > COUNT(1) __C0_1 > FROM PUBLIC.PERSON P__Z0 [90016-195] > -- This message was sent by Atlassian JIRA (v7.6.3#76005)