[SPARK-3211] .take() is OOM-prone when there are empty partitions

• Type: Bug
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 1.0.2
• Fix Version/s: 1.1.1, 1.2.0
• Component/s: Spark Core
• Target Version/s: 1.1.1, 1.2.0
• Assignee: Andrew Ash
• Reporter: Andrew Ash
• Created: 25/Aug/14 19:16
• Updated: 07/Sep/14 06:07
• Resolved: 06/Sep/14 01:52

Description

Filed on dev@ on 22 August by pnepywoda:

On line 777
https://github.com/apache/spark/commit/42571d30d0d518e69eecf468075e4c5a823a2ae8#diff-1d55e54678eff2076263f2fe36150c17R771
the logic for take() reads ALL partitions if the first one (or first k) are
empty. This has actually lead to OOMs when we had many partitions
(thousands) and unfortunately the first one was empty.

Wouldn't a better implementation strategy be

numPartsToTry = partsScanned * 2

instead of

numPartsToTry = totalParts - 1

(this doubling is similar to most memory allocation strategies)

Thanks!

• Paul

Issue Links

• links to: [Github] Pull Request #2117 (ash211)