[IMPALA-5612] Join inversion should avoid reducing the degree of parallelism

• Type: Improvement
• Status: Resolved
• Priority: Critical
• Resolution: Fixed
• Affects Version/s: Impala 2.6.0, Impala 2.7.0, Impala 2.8.0, Impala 2.9.0
• Fix Version/s: Impala 2.10.0
• Component/s: Frontend
• Labels: performance planner ramp-up
• Target Version: Impala 2.10.0
• Epic Color: ghx-label-2
• Assignee: Tim Armstrong
• Reporter: Alexander Behm
• Created: 30/Jun/17 23:04
• Updated: 18/Sep/20 23:03
• Resolved: 22/Aug/17 20:17

Description

The degree of inter-node parallelism for a join is determined by its left input, so when inverting a join the planner should be mindful of how the inversion affects parallelism.

For example, the left join input may have been reduced by joining with several dimension tables so much that it becomes smaller than the right hand-side (another small dimension table). By inverting such a join the degree of parallelism may be reduced to one or very few nodes, based on how many nodes the right-hand size is executed on.

Issue Links

• relates to: IMPALA-10179 After inverting a join's inputs the join's parallelism does not get reset