[CARBONDATA-3345] Presto query in Carbondata-streaming failed

• Type: Bug
• Status: Resolved
• Priority: Blocker
• Resolution: Fixed
• Affects Version/s: 1.5.2
• Component/s: presto-integration
• Labels: features
• Assignee: Unassigned
• Reporter: zjy
• Created: 08/Apr/19 17:58
• Updated: 07/May/19 06:58
• Resolved: 07/May/19 06:58

Description

I use streaming for saving switch' s syslog, here' s my table ddl

CREATE TABLE IF NOT EXISTS syslog(id LONG, device_id LONG, ip STRING, message STRING, level SHORT, message_type CHAR(1), port_index INT, area_id LONG, createdon TIMESTAMP) STORED AS carbondata TBLPROPERTIES ('INVERTED_INDEX'='device_id,level,area_id','SORT_COLUMNS'='device_id,level,area_id,id','streaming'='true')

 

Here' s a record example

 I mock ten thousand switchs，each one produced a record every 10 seconds on a day.

On the early time, the presto' s query is normally.

However, when the data growing, the presto query result is diffent from the spark sql.

 

 

 

 

 

 

 

I am looking forward to resolving it, thanks !

Issue Links

• links to: GitHub Pull Request #3186