[DRILL-4530] Improve metadata cache performance for queries with single partition

• Type: Improvement
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 1.6.0
• Fix Version/s: 1.8.0
• Component/s: Query Planning & Optimization
• Assignee: Aman Sinha
• Reporter: Aman Sinha
• Reviewer: Rahul Kumar Challapalli
• Created: 23/Mar/16 00:51
• Updated: 14/Apr/21 05:57
• Resolved: 19/Jul/16 15:20

Description

Consider two types of queries which are run with Parquet metadata caching:

query 1:
SELECT col FROM  `A/B/C`;

query 2:
SELECT col FROM `A` WHERE dir0 = 'B' AND dir1 = 'C';

For a certain dataset, the query1 elapsed time is 1 sec whereas query2 elapsed time is 9 sec even though both are accessing the same amount of data. The user expectation is that they should perform roughly the same. The main difference comes from reading the bigger metadata cache file at the root level 'A' for query2 and then applying the partitioning filter. query1 reads a much smaller metadata cache file at the subdirectory level.

Issue Links

• Dependent: ZOOKEEPER-704 GSoC 2010: Read-Only Mode
• links to: GitHub Pull Request #519