Github user harishreedharan commented on a diff in the pull request:
https://github.com/apache/spark/pull/2065#discussion_r16727964
--- Diff:
external/flume-sink/src/main/scala/org/apache/spark/streaming/flume/sink/SparkAvroCallbackHandler.scala
---
@@ -62,19 +70,24 @@ private[flume] class SparkAvroCallbackHandler(val
threads: Int, val channel: Cha
*/
override def getEventBatch(n: Int): EventBatch = {
logDebug("Got getEventBatch call from Spark.")
- val sequenceNumber = seqBase + seqCounter.incrementAndGet()
- val processor = new TransactionProcessor(channel, sequenceNumber,
- n, transactionTimeout, backOffInterval, this)
- transactionExecutorOpt.foreach(executor => {
- executor.submit(processor)
- })
- // Wait until a batch is available - will be an error if error message
is non-empty
- val batch = processor.getEventBatch
- if (!SparkSinkUtils.isErrorBatch(batch)) {
- processorMap.put(sequenceNumber.toString, processor)
- logDebug("Sending event batch with sequence number: " +
sequenceNumber)
+ if (stopped) {
+ new EventBatch("Spark sink has been stopped!", "",
java.util.Collections.emptyList())
--- End diff --
Either is fine. Empty batch might be smaller in terms of the amount of data
transferred over the network - since throwing an exception will cause Avro to
serialize the entire exception and send it.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
