[SPARK-35754] Put blocks only on disk while migrating RDD cached data

• Type: Sub-task
• Status: In Progress
• Priority: Major
• Resolution: Unresolved
• Affects Version/s: 3.1.2
• Component/s: Spark Core
• Assignee: Unassigned
• Reporter: abhishek kumar tiwari
• Created: 14/Jun/21 09:32
• Updated: 14/Jun/21 10:23

Description

While migrating cached data we might drop already cached blocks if enough memory is not available on the peer block manager.
We should add a conf to put migrating blocks only on disk

Issue Links

• links to: [Github] Pull Request #32902 (q2w)
• links to: [Github] Pull Request #32902 (q2w)