[SPARK-53882][CONNECT] Add optional sqlstate to sparkthrowables #52696
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This change adds an optional sqlState parameter to the constructors of all SparkThrowable implementations. The getSqlState method is overridden to check the provided property first before falling back to the errorClassReader. Implementation: - Added optional sqlState parameter to all exception constructors - For Scala: sqlState: Option[String] = None - For Java: String sqlState with null as default - Override getSqlState() to prioritize the custom property: * Scala: sqlState.getOrElse(super.getSqlState) * Java: sqlState != null ? sqlState : SparkThrowable.super.getSqlState() Modified Files: - SparkException.scala: 18 exception classes updated - StreamingQueryException.scala - SqlScriptingException.scala - SparkOutOfMemoryError.java - KafkaExceptions.scala: 2 exception classes - SparkFileAlreadyExistsException.scala Test Coverage: - Added 6 tests to SparkThrowableSuite covering Scala and Java exceptions - Created KafkaExceptionsSuite with 3 tests for Kafka-specific exceptions - Tests verify custom SQL state takes precedence over errorClassReader - Tests verify fallback to errorClassReader when no custom state provided 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude <[email protected]>
garlandz-db
changed the title
hvanhovell
reviewed
Nov 11, 2025
|
|
||
| // Portable error identifier across SQL engines | ||
| // If null, error class or SQLSTATE is not set | ||
| // Deprecated: Override this method to provide explicit SQL state instead of relying on error class reader |
Contributor
There was a problem hiding this comment.
Why would this be deprecated?
hvanhovell
reviewed
Nov 11, 2025
Contributor
hvanhovell
left a comment
hvanhovell
left a comment
There was a problem hiding this comment.
One small question. Otherwise LGTM.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
What changes were proposed in this pull request?
Spark connect client cannot properly read the error details when the server has a new error. This is because it relies on an internal ErrorClassReader which pulls from a local json file that contains the error details at that time. However this json file can become out of date when the server throws a new error.
This PR ensures all exceptions will contain an optional sql state property and override the getSqlState to method to first check if the sql state is defined, otherwise use the original functionality (i.e. read from json file). To ensure backward compatibility, all sql states are initially None.
We will also make a client side change in #52589 so that when client receives an exception, it will read the sql state first when reconstructing the exception and populate that exceptions field. Now older clients can carry sql state information for newer servers!
Why are the changes needed?
Does this PR introduce any user-facing change?
How was this patch tested?
Was this patch authored or co-authored using generative AI tooling?