Can't connect to Mongo DB via Spark

Question

I'm trying to read data from Mongo DB through an Apache Spark master.

I'm using 3 machines for this:

• M1 - with a Mongo DB instance on it
• M2 - with a Spark Master, with Mongo connector, running on it
• M3 - with a python application that connects to M2's Spark master

The application(M3) is getting a connection to the spark master like this:

_sparkSession = SparkSession.builder.master(masterPath).appName(appName)\
.config("spark.mongodb.input.uri", "mongodb://10.0.3.150/db1.data.coll")\
.config("spark.mongodb.output.uri", "mongodb://10.0.3.150/db1.data.coll").getOrCreate()

The application(M3) is trying to read data from the DB:

sqlContext = SQLContext(_sparkSession.sparkContext)
        df = sqlContext.read.format("com.mongodb.spark.sql.DefaultSource").option("uri","mongodb://user:[email protected]/db1.data?readPreference=primaryPreferred").load()

but fails with this exception:

    py4j.protocol.Py4JJavaError: An error occurred while calling o56.load.
: java.lang.ClassNotFoundException: Failed to find data source: com.mongodb.spark.sql.DefaultSource. Please find packages at http://spark.apache.org/third-party-projects.html
        at org.apache.spark.sql.execution.datasources.DataSource$.lookupDataSource(DataSource.scala:594)
        at org.apache.spark.sql.execution.datasources.DataSource.providingClass$lzycompute(DataSource.scala:86)
        at org.apache.spark.sql.execution.datasources.DataSource.providingClass(DataSource.scala:86)
        at org.apache.spark.sql.execution.datasources.DataSource.resolveRelation(DataSource.scala:325)
        at org.apache.spark.sql.DataFrameReader.load(DataFrameReader.scala:152)
        at org.apache.spark.sql.DataFrameReader.load(DataFrameReader.scala:125)
        at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
        at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62)
        at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
        at java.lang.reflect.Method.invoke(Method.java:498)
        at py4j.reflection.MethodInvoker.invoke(MethodInvoker.java:244)
        at py4j.reflection.ReflectionEngine.invoke(ReflectionEngine.java:357)
        at py4j.Gateway.invoke(Gateway.java:280)
        at py4j.commands.AbstractCommand.invokeMethod(AbstractCommand.java:132)
        at py4j.commands.CallCommand.execute(CallCommand.java:79)
        at py4j.GatewayConnection.run(GatewayConnection.java:214)
        at java.lang.Thread.run(Thread.java:748)
Caused by: java.lang.ClassNotFoundException: com.mongodb.spark.sql.DefaultSource.DefaultSource
        at java.net.URLClassLoader.findClass(URLClassLoader.java:381)
        at java.lang.ClassLoader.loadClass(ClassLoader.java:424)
        at java.lang.ClassLoader.loadClass(ClassLoader.java:357)
        at org.apache.spark.sql.execution.datasources.DataSource$$anonfun$25$$anonfun$apply$13.apply(DataSource.scala:579)
        at org.apache.spark.sql.execution.datasources.DataSource$$anonfun$25$$anonfun$apply$13.apply(DataSource.scala:579)
        at scala.util.Try$.apply(Try.scala:192)
        at org.apache.spark.sql.execution.datasources.DataSource$$anonfun$25.apply(DataSource.scala:579)
        at org.apache.spark.sql.execution.datasources.DataSource$$anonfun$25.apply(DataSource.scala:579)
        at scala.util.Try.orElse(Try.scala:84)
        at org.apache.spark.sql.execution.datasources.DataSource$.lookupDataSource(DataSource.scala:579)
        ... 16 more

Solution 1

Spark can't find the com.mongodb.spark.sql.DefaultSource package, hence the error message.

Everything, else looks good just need to include the Mongo Spark package:

> $SPARK_HOME/bin/pyspark --packages org.mongodb.spark:mongo-spark-connector_2.11:2.2.0

Or ensure that the jar file is on the correct path.

Make sure you check the version of the Mongo-Spark package required for your version of Spark: https://spark-packages.org/package/mongodb/mongo-spark

Solution 2

I am a pyspark user, here is what my code looks like, and it works:

MongoDB connection configuration in pyspark

# For spark version < 3.0
from pyspark.sql import SparkSession
spark = SparkSession\
    .builder\
    .master('local')\
    .config('spark.mongodb.input.uri', 'mongodb://user:[email protected]:27017/database01.data.coll')\
    .config('spark.mongodb.output.uri', 'mongodb://user:[email protected]:27017/database01.data.coll')\
    .config('spark.jars.packages', 'org.mongodb.spark:mongo-spark-connector_2.11:2.3.1')\
    .getOrCreate()

# For spark version >= 3.0
from pyspark.sql import SparkSession
spark = SparkSession\
    .builder\
    .master('local')\
    .config('spark.mongodb.input.uri', 'mongodb://user:[email protected]:27017/database01.coll')\
    .config('spark.mongodb.output.uri', 'mongodb://user:[email protected]:27017/database01.coll')\
    .config('spark.jars.packages', 'org.mongodb.spark:mongo-spark-connector_2.12:3.0.1')\
    .getOrCreate()

Read from MongoDB:

df01 = spark.read\
    .format("com.mongodb.spark.sql.DefaultSource")\
    .option("database","database01")\
    .option("collection", "collection01")\
    .load()

Write to MongoDB:

df01.write.format("com.mongodb.spark.sql.DefaultSource")\
    .mode("overwrite")\
    .option("database","database01")\
    .option("collection", "collection02")\
    .save()

Solution 3

I have had a quite hard time configuring the Spark connection to CosmosDB (API MongoDB), so I decided to post the code that worked for me as a contribution.

I used Spark 2.4.0 through a Databricks notebook.

from pyspark.sql import SparkSession

# Connect to CosmosDB to write on the collection
userName = "userName"
primaryKey = "myReadAndWritePrimaryKey"
host = "ipAddress"
port = "10255"
database = "dbName"
collection = "collectionName"

# Structure the connection
connectionString = "mongodb://{0}:{1}@{2}:{3}/{4}.{5}?ssl=true&replicaSet=globaldb".format(userName, primaryKey, host, port, database, collection)

spark = SparkSession\
    .builder\
    .config('spark.mongodb.input.uri', connectionString)\
    .config('spark.mongodb.output.uri', connectionString)\
    .config('spark.jars.packages', 'org.mongodb.spark:mongo-spark-connector_2.11:2.3.1')\
    .getOrCreate()

# Reading from CosmosDB
df = spark.read\
    .format("com.mongodb.spark.sql.DefaultSource")\
    .option("uri", connectionString)\
    .option("database", database)\
    .option("collection", collection)\
    .load()

# Writing on CosmosDB (Appending new information without replacing documents)
dfToAppendOnCosmosDB.write.format("com.mongodb.spark.sql.DefaultSource")\
    .mode("append")\
    .option("uri", connectionString)\
    .option("replaceDocument", False)\
    .option("maxBatchSize", 100)\
    .option("database", database)\
    .option("collection", collection)\
    .save()

I found the options to configure the connector at the link.

Comments

[Ran P]

Thank you for your answer. I specified that I ran the app through a remote Python application, and not by the PySpark shell. So, as a noob python developer, I ask again, how do I run my application with the connector package. Or do I need to run the spark master with the package?

[Ross]

Please update the question with more information on how you submit your spark jobs and I'll look to update my answer.

[Ran P]

I changed the way I use the spark master. I initiate the Spark master and its slaves. After that I run the spark-submit with the mongo-spark-connector package and the python script. Guess that's the recommended way. Thanx all

[Farhat Nawaz]

@Ross I have the same issue and can't seem to resolve it. any ideas?

[Will]

When I ran into this, I didn't have the mongodb-spark-connector_2.11-2.2.3.jar in my $SPARK_HOME/jars (e.g. /usr/local/spark-2.2.2-bin-hadoop2.7/jars).

[Gaboik1]

This should be the accepted answer. The spark.jars.packages option is documented at spark.apache.org/docs/2.1.0/…