erenavsarogullari commented on a change in pull request #34980:
URL: https://github.com/apache/spark/pull/34980#discussion_r776122976
##########
File path: core/src/main/scala/org/apache/spark/storage/BlockManager.scala
##########
@@ -926,30 +933,48 @@ private[spark] class BlockManager(
})
Some(new BlockResult(ci, DataReadMethod.Memory, info.size))
} else if (level.useDisk && diskStore.contains(blockId)) {
- val diskData = diskStore.getBytes(blockId)
- val iterToReturn: Iterator[Any] = {
- if (level.deserialized) {
- val diskValues = serializerManager.dataDeserializeStream(
- blockId,
- diskData.toInputStream())(info.classTag)
- maybeCacheDiskValuesInMemory(info, blockId, level, diskValues)
- } else {
- val stream = maybeCacheDiskBytesInMemory(info, blockId, level,
diskData)
- .map { _.toInputStream(dispose = false) }
- .getOrElse { diskData.toInputStream() }
- serializerManager.dataDeserializeStream(blockId,
stream)(info.classTag)
+ try {
+ val diskData = diskStore.getBytes(blockId)
+ val iterToReturn: Iterator[Any] = {
+ if (level.deserialized) {
+ val diskValues = serializerManager.dataDeserializeStream(
+ blockId,
+ diskData.toInputStream())(info.classTag)
+ maybeCacheDiskValuesInMemory(info, blockId, level, diskValues)
+ } else {
+ val stream = maybeCacheDiskBytesInMemory(info, blockId, level,
diskData)
+ .map { _.toInputStream(dispose = false) }
+ .getOrElse { diskData.toInputStream() }
+ serializerManager.dataDeserializeStream(blockId,
stream)(info.classTag)
+ }
}
+ val ci = CompletionIterator[Any, Iterator[Any]](iterToReturn, {
+ releaseLockAndDispose(blockId, diskData, taskContext)
+ })
+ Some(new BlockResult(ci, DataReadMethod.Disk, info.size))
+ } catch {
+ case ex: KryoException if ex.getCause.isInstanceOf[IOException] =>
+ // We need to have clear error message to catch environmental
problems easily.
+ // Further details:
https://issues.apache.org/jira/browse/SPARK-37710
+ processKryoException(ex, blockId)
+ throw ex
}
- val ci = CompletionIterator[Any, Iterator[Any]](iterToReturn, {
- releaseLockAndDispose(blockId, diskData, taskContext)
- })
- Some(new BlockResult(ci, DataReadMethod.Disk, info.size))
} else {
handleLocalReadFailure(blockId)
}
}
}
+ private def processKryoException(ex: KryoException, blockId: BlockId): Unit
= {
+ var errorMessage = s"${ex.getMessage}. Please check if environment status
is healthy for " +
+ s"disk corruption, network failure (etc). ${blockManagerId.toString} -
blockName: $blockId"
Review comment:
In this case, `IOException` is wrapped by `KryoException`. This means it
occurs at `Kryo deserialization` level when accessing to I/O via disk/network.
In our use-case, this occurred due to `disk corruption` (with error message:
`java.io.IOException: Input/output error`). So, would like to highlight `disk
corruption` here as a real-life example for future references.
`etc` (in the log) already aims to explain as there may be different error
cases for `IOException`. This kind of environmental problems hard to catch and
they require correlation with infrastructure alerts(e.g: reporting disk
corruption) if the pipeline has so as much as possible, providing hints will
help for future RCAs.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
