[jira] [Updated] (ATLAS-347) Atlas search APIs should allow pagination of results
[ https://issues.apache.org/jira/browse/ATLAS-347?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Shwetha G S updated ATLAS-347: -- Attachment: ATLAS-347-v2.patch orderby doesn't require fully qualified field name anymore. For example, 'hive_table orderby name' resolves name to 'hive_table orderby Asset.name' limit and offset in API and queries are merged - final limit=min(API limit, query limit), final offset=(API offset + query offset) limit and offset params in API are optional and have 100, and 0 as defaults respectively Pagination of full text search results will be handled in another jira DSL query supported multiple query source, for example 'hive_table as t where name = 'xx', hive_db as d where name = 'yy' select t.name, d.name'. There can be single select in the query. But the QueryParser had wrong syntx of supporting multiple selects, fixed that. So, DSL query can contain multiple query sources with where clause, but can contain single select, orderby and limit orderby using alias doesn't work - hive_table select name as n orderby n. But hive_table select name as n orderby 'n' works. Created ATLAS-1007 Added/modified tests > Atlas search APIs should allow pagination of results > > > Key: ATLAS-347 > URL: https://issues.apache.org/jira/browse/ATLAS-347 > Project: Atlas > Issue Type: Improvement >Reporter: Hemanth Yamijala >Assignee: Shwetha G S >Priority: Critical > Fix For: 0.8-incubating > > Attachments: ATLAS-347-v2.patch, ATLAS-347.patch > > > Currently, we can search for all entities of type like a hive table or hive > column (query string is hive_table or hive_column). The API returns all > results at one shot in this case. With a sufficiently large number of > entities, this would be a huge list and a lot of data flowing from server to > UI. To scale this better, we should have a way of specifying to get a page of > results at a time in the API. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Updated] (ATLAS-347) Atlas search APIs should allow pagination of results
[ https://issues.apache.org/jira/browse/ATLAS-347?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Madhan Neethiraj updated ATLAS-347: --- Priority: Critical (was: Major) > Atlas search APIs should allow pagination of results > > > Key: ATLAS-347 > URL: https://issues.apache.org/jira/browse/ATLAS-347 > Project: Atlas > Issue Type: Improvement >Reporter: Hemanth Yamijala >Assignee: Shwetha G S >Priority: Critical > > Currently, we can search for all entities of type like a hive table or hive > column (query string is hive_table or hive_column). The API returns all > results at one shot in this case. With a sufficiently large number of > entities, this would be a huge list and a lot of data flowing from server to > UI. To scale this better, we should have a way of specifying to get a page of > results at a time in the API. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Updated] (ATLAS-347) Atlas search APIs should allow pagination of results
[ https://issues.apache.org/jira/browse/ATLAS-347?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Madhan Neethiraj updated ATLAS-347: --- Assignee: Shwetha G S > Atlas search APIs should allow pagination of results > > > Key: ATLAS-347 > URL: https://issues.apache.org/jira/browse/ATLAS-347 > Project: Atlas > Issue Type: Improvement >Reporter: Hemanth Yamijala >Assignee: Shwetha G S > > Currently, we can search for all entities of type like a hive table or hive > column (query string is hive_table or hive_column). The API returns all > results at one shot in this case. With a sufficiently large number of > entities, this would be a huge list and a lot of data flowing from server to > UI. To scale this better, we should have a way of specifying to get a page of > results at a time in the API. -- This message was sent by Atlassian JIRA (v6.3.4#6332)