Ashutosh Mestry created ATLAS-2447: -------------------------------------- Summary: Advanced Search DSL: Improve Select Queries Using De-duplication During Query Formation Key: ATLAS-2447 URL: https://issues.apache.org/jira/browse/ATLAS-2447 Project: Atlas Issue Type: Bug Components: atlas-core Affects Versions: trunk Reporter: Ashutosh Mestry Assignee: Ashutosh Mestry Fix For: trunk
In the provided Atlas instance , there are many 1000+ hive_table entities . Few of them : db2.table1 db1.table1 db1.table2 db3.table2 DSL query : {code:java} typename = hive_table query = name > "table1" {code} listed all the hive_table entities whose name>"table1" , With limit 25 per page , db3.table2 is in 19th page (467th entity) db1.table2 is in 21st page (521th entity) *+Error scenario :+* DSL query : {code:java} typename = hive_db query = where name > "table1" select db {code} limit = 25 , lists only 1 db . Expected that it would result db3 and db1 along with other db's since "table2" is greater than "table1" and "table2" is in db3 and db1. when same query with limit =1000 , lists many results along with db1 and db3. When limit is 25 , "select db" is done on the returned 25 results of "name > table1" and selects the db. Since table2 doesn't appear in the first 25 entities , db3 and db1 didn't result. Attaching the screenshots for limit 10000 and limit 25. -- This message was sent by Atlassian JIRA (v7.6.3#76005)