org.apache.spark.sql.AnalysisException: 'write' can not be called on streaming Dataset/DataFrame - apache-spark

Question

I'm trying to write a Spark Structured Streaming (2.3) dataset to ScyllaDB (Cassandra).
My code to write the dataset:
def saveStreamSinkProvider(ds: Dataset[InvoiceItemKafka]) = {
ds
.writeStream
.format("cassandra.ScyllaSinkProvider")
.outputMode(OutputMode.Append)
.queryName("KafkaToCassandraStreamSinkProvider")
.options(
Map(
"keyspace" -> namespace,
"table" -> StreamProviderTableSink,
"checkpointLocation" -> "/tmp/checkpoints"
)
)
.start()
}
My ScyllaDB Streaming Sinks:
class ScyllaSinkProvider extends StreamSinkProvider {
override def createSink(sqlContext: SQLContext,
parameters: Map[String, String],
partitionColumns: Seq[String],
outputMode: OutputMode): ScyllaSink =
new ScyllaSink(parameters)
}
class ScyllaSink(parameters: Map[String, String]) extends Sink {
override def addBatch(batchId: Long, data: DataFrame): Unit =
data.write
.cassandraFormat(
parameters("table"),
parameters("keyspace")
//parameters("cluster")
)
.mode(SaveMode.Append)
.save()
}
However, when I run this code, I receive an exception:
...
[error] +- StreamingExecutionRelation KafkaSource[Subscribe[transactions_load]], [key#7, value#8, topic#9, partition#10, offset#11L, timestamp#12, timestampType#13]
[error] at org.apache.spark.sql.execution.streaming.StreamExecution.org$apache$spark$sql$execution$streaming$StreamExecution$$runStream(StreamExecution.scala:295)
[error] at org.apache.spark.sql.execution.streaming.StreamExecution$$anon$1.run(StreamExecution.scala:189)
[error] Caused by: org.apache.spark.sql.AnalysisException: 'write' can not be called on streaming Dataset/DataFrame;
[error] at org.apache.spark.sql.catalyst.analysis.package$AnalysisErrorAt.failAnalysis(package.scala:42)
[error] at org.apache.spark.sql.Dataset.write(Dataset.scala:3103)
[error] at cassandra.ScyllaSink.addBatch(CassandraDriver.scala:113)
[error] at org.apache.spark.sql.execution.streaming.MicroBatchExecution$$anonfun$org$apache$spark$sql$execution$streaming$MicroBatchExecution$$runBatch$3$$anonfun$apply$16.apply(MicroBatchExecution.scala:477)
...
I have seen a similar question, but that is for CosmosDB - Spark CosmosDB Sink: org.apache.spark.sql.AnalysisException: 'write' can not be called on streaming Dataset/DataFrame

Answer 1

You could convert it to an RDD first and then write:
class ScyllaSink(parameters: Map[String, String]) extends Sink {
override def addBatch(batchId: Long, data: DataFrame): Unit = synchronized {
val schema = data.schema
// this ensures that the same query plan will be used
val rdd: RDD[Row] = df.queryExecution.toRdd.mapPartitions { rows =>
val converter = CatalystTypeConverters.createToScalaConverter(schema)
rows.map(converter(_).asInstanceOf[Row])
}
// write the RDD to Cassandra
}
}