[
https://issues.apache.org/jira/browse/DRILL-1319?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Jacques Nadeau updated DRILL-1319:
----------------------------------
Fix Version/s: 0.6.0
> TPCDS Query 65 takes longer than usual to execute
> -------------------------------------------------
>
> Key: DRILL-1319
> URL: https://issues.apache.org/jira/browse/DRILL-1319
> Project: Apache Drill
> Issue Type: Bug
> Components: Query Planning & Optimization
> Reporter: Abhishek Girish
> Fix For: 0.6.0
>
> Attachments: ExplanQ65.rtf, query65.rtf
>
>
> Suite: TPCDS (using Impala TPCDS-KIT)
> ScaleFactor: 3000
> Query: 65 (attached)
> Query Plan:
> +------------+------------+
> | text | json |
> +------------+------------+
> | 00-00 Screen
> 00-01 Project(s_store_name=[$0], i_item_desc=[$1], revenue=[$2],
> i_current_price=[$3], i_wholesale_cost=[$4], i_brand=[$5])
> 00-02 SelectionVectorRemover
> 00-03 Limit(fetch=[100])
> 00-04 SelectionVectorRemover
> 00-05 |
> +------------+------------+
> 1 row selected (9.864 seconds)
> Run times:
> 1. 627 seconds
> 2. 1516 seconds
> 3. 1316 seconds
> Cluster details:
> Number of nodes: 4 (3 zookeepers and 3 drillbits)
> Cluster RAM: 128GB each node - 512 total
> Cluster disk space must not be a constraint.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
