I am practicing with Databricks. In sample notebooks, I have seen different use of writeStream with or without .start() method. I have a few questions in this regard.
Samples are below:
Without .start():
spark.readStream
.format("cloudFiles")
.option("cloudFiles.format", source_format)
.option("cloudFiles.schemaLocation", checkpoint_directory)
.load(data_source)
.writeStream
.option("checkpointLocation", checkpoint_directory)
.option("mergeSchema", "true")
.table(table_name)
With .start():
(myDF
.writeStream
.format("delta")
.option("checkpointLocation", checkpointPath)
.outputMode("append")
.start(path)
)
With .start():
query = (streaming_df.writeStream
.foreachBatch(streaming_merge.upsert_to_delta)
.outputMode("update")
.option("checkpointLocation", checkpoint_directory)
.trigger(availableNow=True)
.start())
query.awaitTermination()
Q1) I didn't understand where should / shouldn't use .start() method. I appreciate it if you could guide me on this.
Q2) If I don't pass path to the start(), where the data files will be written?
Related
I am not being to stream my data on multiple hdfs location , which is filtered by key. So below code is not working. Please help me to find the correct way to write this code
val ER_stream_V1 = spark
.readStream
.format("kafka")
.option("kafka.bootstrap.servers", configManager.getString("Kafka.Server"))
.option("subscribe", "Topic1")
.option("startingOffsets", "latest")
.option("failOnDataLoss", "false")
.load()
val ER_stream_V2 = spark
.readStream
.format("kafka")
.option("kafka.bootstrap.servers", configManager.getString("Kafka.Server"))
.option("subscribe", "Topic1")
.option("startingOffsets", "latest")
.option("failOnDataLoss", "false")
.load()
ER_stream_V1.toDF()
.select(col("key"), col("value").cast("string"))
.filter(col("key")==="Value1")
.select(functions.from_json(col("value").cast("string"), Value1Schema.schemaExecution).as("value")).select("value.*")
.writeStream
.format("orc")
.option("metastoreUri", configManager.getString("spark.datasource.hive.warehouse.metastoreUri"))
.option("checkpointLocation", "/tmp/teststreaming/execution/checkpoint2005")
.option("path", "/tmp/test/value1")
.trigger(Trigger.ProcessingTime("5 Seconds"))
.partitionBy("jobid")
.start()
ER_stream_V2.toDF()
.select(col("key"), col("value").cast("string"))
.filter(col("key")==="Value2")
.select(functions.from_json(col("value").cast("string"), Value2Schema.schemaJobParameters).as("value"))
.select("value.*")
.writeStream
.format("orc")
.option("metastoreUri", configManager.getString("spark.datasource.hive.warehouse.metastoreUri"))
.option("checkpointLocation", "/tmp/teststreaming/jobparameters/checkpoint2006")
.option("path", "/tmp/test/value2")
.trigger(Trigger.ProcessingTime("5 Seconds"))
.partitionBy("jobid")
.start()
You should not need two readers. Create one and filter twice. You might also want to consider startingOffsets as earliest to read existing topic data
For example.
val ER_stream = spark
.readStream
.format("kafka")
.option("kafka.bootstrap.servers", configManager.getString("Kafka.Server"))
.option("subscribe", "Topic1")
.option("startingOffsets", "latest") // maybe change?
.option("failOnDataLoss", "false")
.load()
.toDF()
.select(col("key").cast("string").as("key"), col("value").cast("string"))
val value1Stream = ER_stream
.filter(col("key") === "Value1")
.select(functions.from_json(col("value"), Value1Schema.schemaExecution).as("value"))
.select("value.*")
val value2Stream = ER_stream
.filter(col("key") === "Value2")
.select(functions.from_json(col("value"), Value2Schema.schemaJobParameters).as("value"))
.select("value.*")
value1Stream.writeStream.format("orc")
...
.start()
value2Stream.writeStream.format("orc")
...
.start()
I am trying to write stream dataset to snowflake as indicated below
val query = expandedDF.writeStream
.trigger(Trigger.ProcessingTime("30 seconds"))
.foreachBatch{(batchDF:DataFrame, batchID: Long) =>
batchDF.write
.format(SNOWFLAKE_SOURCE_NAME)
.options(sfOptions)
.option("dbtable", "TWITTER")
.mode(SaveMode.Append)
.save()
}
.outputMode("update")
.start()
query.awaitTermination()
However, whenever this command is executed I run into the error
overloaded method foreachBatch with alternatives
anyone here knows how I can resolve this?
I am trying to write to streams parallel, but just the last one is working.
Here is my code.
val trump_topic = filteredQueryTrump.select($"ID".as("key"), $"title".as("value"))
.writeStream.format("kafka")
.trigger(Trigger.ProcessingTime("10 seconds"))
.option("kafka.security.protocol", "SASL_SSL")
.option("kafka.sasl.mechanism", "SCRAM-SHA-256")
.option("kafka.sasl.jaas.config", """org.apache.kafka.common.security.scram.ScramLoginModule required username="" password="";""")
.option("kafka.bootstrap.servers", "rocket-01.srvs.cloudkafka.com:9094")
.option("topic", "trump")
.option("checkpointLocation", "/tmp").start()
val biden_topic = filteredQueryBiden.select($"ID".as("key"), $"title".as("value"))
.writeStream.format("kafka")
.trigger(Trigger.ProcessingTime("4 seconds"))
.option("kafka.security.protocol", "SASL_SSL")
.option("kafka.sasl.mechanism", "SCRAM-SHA-256")
.option("kafka.sasl.jaas.config", """org.apache.kafka.common.security.scram.ScramLoginModule required username="" password="";""")
.option("kafka.bootstrap.servers", "rocket-01.srvs.cloudkafka.com:9094")
.option("topic", "biden")
.option("checkpointLocation", "/tmp").start()
I tried spark.streams.awaitAnyTermination() but it didn't help either.
You need to have two different checkpoint locations for each output stream.
Also, do add awaitAnyTermination as you already have tried out to ensure the jobs are running continously.
spark.streams.awaitAnyTermination()
As an alternative you could combine those two queries into one by doing
val combinedDf = filteredQueryTrump.select($"ID".as("key"), $"title".as("value"), lit("trump").as("topic"))
.union(filteredQueryBiden.select($"ID".as("key"), $"title".as("value"), lit("biden").as("topic")))
The content of the column topic will be interpreted by the spark-sql-kafka-0-10 library to decide into which topic the row should be written to. Theoretically, you can use that technique to write to multiple Kafka topics within one stream. When doing this it is important to not use topic option when calling writeStream. However, in that way you will not be able to have two different Trigger durations but only one for the overall stream.
An example you code could look like this:
val outputQuery = combinedDf.select($"key", $"value", $"topic")
.writeStream
.format("kafka")
.trigger(Trigger.ProcessingTime("4 seconds"))
.option("kafka.security.protocol", "SASL_SSL")
.option("kafka.sasl.mechanism", "SCRAM-SHA-256")
.option("kafka.sasl.jaas.config", """org.apache.kafka.common.security.scram.ScramLoginModule required username="" password="";""")
.option("kafka.bootstrap.servers", "rocket-01.srvs.cloudkafka.com:9094")
.option("checkpointLocation", "/tmp")
outputQuery.start()
outputQuery.awaitTermination()
I have a Spark Structured Streaming:
val df = spark
.readStream
.format("kafka")
.option("kafka.bootstrap.servers", "localhost:9092")
.option("startingOffsets", "earliest")
.option("endingOffsets", "latest")
.option("subscribe", "topic")
.load()
I want to write data to FileSystem using DataStreamWriter,
val query = df
.writeStream
.outputMode("append")
.format("parquet")
.start("data")
But zero files are getting created in data folder. Only _spark_metadata is getting created.
However, I can see the data on console when format is console:
val query = df
.writeStream
.outputMode("append")
.format("console")
.start()
+--------------------+------------------+------------------+
| time| col1| col2|
+--------------------+------------------+------------------+
|49368-05-11 20:42...|0.9166470338147503|0.5576946794171861|
+--------------------+------------------+------------------+
I cannot understand the reason behind it.
Spark - 2.1.0
I had a similar problem but for different reasons, posting here in case someone has the same issue. When writing your output stream to file in append mode with watermarking, structured streaming has an interesting behavior where it won't actually write any data until a time bucket is older than the watermark time. If you're testing structured streaming and have an hour long water mark, you won't see any output for at least an hour.
I resolved this issue. Actually when I tried to run the Structured Streaming on spark-shell, then it gave an error that endingOffsets are not valid in streaming queries, i.e.,:
val df = spark
.readStream
.format("kafka")
.option("kafka.bootstrap.servers", "localhost:9092")
.option("startingOffsets", "earliest")
.option("endingOffsets", "latest")
.option("subscribe", "topic")
.load()
java.lang.IllegalArgumentException: ending offset not valid in streaming queries
at org.apache.spark.sql.kafka010.KafkaSourceProvider$$anonfun$validateStreamOptions$1.apply(KafkaSourceProvider.scala:374)
at org.apache.spark.sql.kafka010.KafkaSourceProvider$$anonfun$validateStreamOptions$1.apply(KafkaSourceProvider.scala:373)
at scala.Option.map(Option.scala:146)
at org.apache.spark.sql.kafka010.KafkaSourceProvider.validateStreamOptions(KafkaSourceProvider.scala:373)
at org.apache.spark.sql.kafka010.KafkaSourceProvider.sourceSchema(KafkaSourceProvider.scala:60)
at org.apache.spark.sql.execution.datasources.DataSource.sourceSchema(DataSource.scala:199)
at org.apache.spark.sql.execution.datasources.DataSource.sourceInfo$lzycompute(DataSource.scala:87)
at org.apache.spark.sql.execution.datasources.DataSource.sourceInfo(DataSource.scala:87)
at org.apache.spark.sql.execution.streaming.StreamingRelation$.apply(StreamingRelation.scala:30)
at org.apache.spark.sql.streaming.DataStreamReader.load(DataStreamReader.scala:124)
... 48 elided
So, I removed endingOffsets from streaming query.
val df = spark
.readStream
.format("kafka")
.option("kafka.bootstrap.servers", "localhost:9092")
.option("startingOffsets", "earliest")
.option("subscribe", "topic")
.load()
Then I tried to save streaming queries' result in Parquet files, during which I came to know that - checkpoint location must be specified, i.e.,:
val query = df
.writeStream
.outputMode("append")
.format("parquet")
.start("data")
org.apache.spark.sql.AnalysisException: checkpointLocation must be specified either through option("checkpointLocation", ...) or SparkSession.conf.set("spark.sql.streaming.checkpointLocation", ...);
at org.apache.spark.sql.streaming.StreamingQueryManager$$anonfun$3.apply(StreamingQueryManager.scala:207)
at org.apache.spark.sql.streaming.StreamingQueryManager$$anonfun$3.apply(StreamingQueryManager.scala:204)
at scala.Option.getOrElse(Option.scala:121)
at org.apache.spark.sql.streaming.StreamingQueryManager.createQuery(StreamingQueryManager.scala:203)
at org.apache.spark.sql.streaming.StreamingQueryManager.startQuery(StreamingQueryManager.scala:269)
at org.apache.spark.sql.streaming.DataStreamWriter.start(DataStreamWriter.scala:262)
at org.apache.spark.sql.streaming.DataStreamWriter.start(DataStreamWriter.scala:206)
... 48 elided
So, I added checkPointLocation:
val query = df
.writeStream
.outputMode("append")
.format("parquet")
.option("checkpointLocation", "checkpoint")
.start("data")
After doing these modifications, I was able to save streaming queries' results in Parquet files.
But, it is strange that when I ran the same code via sbt application, it didn't threw any errors, but when I ran the same code via spark-shell these errors were thrown. I think Apache Spark should throw these errors when run via sbt/maven app too. It is seems to be a bug to me !
I'm trying to read the messages from kafka (version 10) in spark and trying to print it.
import spark.implicits._
val spark = SparkSession
.builder
.appName("StructuredNetworkWordCount")
.config("spark.master", "local")
.getOrCreate()
val ds1 = spark.readStream.format("kafka")
.option("kafka.bootstrap.servers", "localhost:9092")
.option("subscribe", "topicA")
.load()
ds1.collect.foreach(println)
ds1.writeStream
.format("console")
.start()
ds1.printSchema()
getting an error Exception in thread "main"
org.apache.spark.sql.AnalysisException: Queries with streaming sources
must be executed with writeStream.start();;
You are branching the query plan: from the same ds1 you are trying to:
ds1.collect.foreach(...)
ds1.writeStream.format(...){...}
But you are only calling .start() on the second branch, leaving the other dangling without a termination, which in turn throws the exception you are getting back.
The solution is to start both branches and await termination.
val ds1 = spark.readStream.format("kafka")
.option("kafka.bootstrap.servers", "localhost:9092")
.option("subscribe", "topicA")
.load()
val query1 = ds1.collect.foreach(println)
.writeStream
.format("console")
.start()
val query2 = ds1.writeStream
.format("console")
.start()
ds1.printSchema()
query1.awaitTermination()
query2.awaitTermination()
I struggled a lot with this issue. I tried each of suggested solution from various blog.
But I my case there are few statement in between calling start() on query and finally at last i was calling awaitTerminate() function that cause this.
Please try in this fashion, It is perfectly working for me.
Working example:
val query = df.writeStream
.outputMode("append")
.format("console")
.start().awaitTermination();
If you write in this way that will cause exception/ error:
val query = df.writeStream
.outputMode("append")
.format("console")
.start()
// some statement
// some statement
query.awaitTermination();
will throw given exception and will close your streaming driver.
i fixed issue by using following code.
val df = session
.readStream
.format("kafka")
.option("kafka.bootstrap.servers", brokers)
.option("subscribe", "streamTest2")
.load();
val query = df.writeStream
.outputMode("append")
.format("console")
.start()
query.awaitTermination()
Kindly remove ds1.collect.foreach(println) and ds1.printSchema() , use outputMode and awaitAnyTermination for background process Waiting until any of the queries on the associated spark.streams has terminated
val spark = SparkSession
.builder
.appName("StructuredNetworkWordCount")
.config("spark.master", "local[*]")
.getOrCreate()
val ds1 = spark.readStream.format("kafka")
.option("kafka.bootstrap.servers", "localhost:9092")
.option("subscribe", "topicA") .load()
val consoleOutput1 = ds1.writeStream
.outputMode("update")
.format("console")
.start()
spark.streams.awaitAnyTermination()
|key|value|topic|partition|offset|
+---+-----+-----+---------+------+
+---+-----+-----+---------+------+
I was able to resolves this issue by following code. In my scenario, I had multiple intermediate Dataframes, which were basically the transformations made on the inputDF.
val query = joinedDF
.writeStream
.format("console")
.option("truncate", "false")
.outputMode(OutputMode.Complete())
.start()
.awaitTermination()
joinedDF is the result of the last transformation performed.