[SPARK-33587] Kill the executor on nested fatal errors

• Type: Improvement
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 3.0.1
• Fix Version/s: 3.1.0
• Component/s: Spark Core
• Assignee: Shixiong Zhu
• Reporter: Shixiong Zhu
• Created: 28/Nov/20 18:18
• Updated: 29/Nov/20 19:57
• Resolved: 29/Nov/20 19:57

Description

Currently we kill the executor when hitting a fatal error. However, if the fatal error is wrapped by another exception, such as

• java.util.concurrent.ExecutionException, com.google.common.util.concurrent.UncheckedExecutionException, com.google.common.util.concurrent.ExecutionError when using Guava cache and java thread pool.
• SparkException thrown from this line: https://github.com/apache/spark/blob/cf98a761de677c733f3c33230e1c63ddb785d5c5/sql/core/src/main/scala/org/apache/spark/sql/execution/datasources/FileFormatWriter.scala#L231

We will still keep the executor running. Fatal errors are usually unrecoverable (such as OutOfMemoryError), some components may be in a broken state when hitting a fatal error. Hence, it's better to detect the nested fatal error as well and kill the executor. Then we can rely on Spark's fault tolerance to recover.

Issue Links

• links to: [Github] Pull Request #30528 (zsxwing)