I am attempting to run a spark job that accesses dynamodb and the old way of instantiating a dynamoDb client has been deprecated and it is now recommended to use the client builder.
Well, this works fine locally, but when I deploy to EMR i'm getting this error:
Exception in thread "main" java.lang.IllegalAccessError: tried to access class com.amazonaws.services.dynamodbv2.AmazonDynamoDBClientConfigurationFactory from class com.amazonaws.services.dynamodbv2.AmazonDynamoDBAsyncClientBuilder
My code that causes this is:
val dynamoDbClient = AmazonDynamoDBAsyncClientBuilder
.standard()
.withRegion(Regions.US_EAST_1)
.build()
my build.sbt contains:
libraryDependencies += "com.amazonaws" % "aws-java-sdk" % "1.11.114"
and my spark-submit command looks like this:
spark-submit --conf spark.eventLog.enabled=false --packages com.typesafe.play:play-json_2.11:2.5.9,com.github.traviscrawford:spark-dynamodb:0.0.6,com.amazonaws:aws-java-sdk:1.11.114 --master yarn --deploy-mode cluster --class Main application.jar
Does anyone have any ideas? Am I overlooking something basic?
Update
I noticed that EMR was running OpenJDK 1.8 and my local system was running Oracle Java 1.8. I changed the EMR cluster to match the java I was running, but there was still no change.
I dont have a perfect answer here but I'm struggling with a similar problem with a fat jar build Spark Driver running on EMR. So I drop my recent tour.
Try to run spark-submit with option -v and look into the logs about class paths and so forth. As I can see EMR is loading an aws-java-sdk as well. Its not clear to me which version of aws-java-sdk EMR is running? EMR release 4.7.0 states "Upgraded the AWS SDK for Java to 1.10.75" (http://docs.aws.amazon.com/emr/latest/ReleaseGuide/emr-whatsnew.html).
Then add another argument --conf spark.driver.userClassPathFirst=true
to load the aws-java-sdk version your driver specifies.
Unfortunately the last step raises yarn errors like: Unable to load YARN support ... (some discussion on that: https://community.cloudera.com/t5/Advanced-Analytics-Apache-Spark/spark-submit-fails-after-setting-userClassPathFirst-to-true/td-p/46778)
Some discussion from the aws-java-sdk github repos: https://github.com/aws/aws-sdk-java/issues/1094
Conclusion: For now use apis of aws-java-sdk version 1.10.75
Related
Need this jar for listening to metrics data for streaming query. The jar works when passed from file system using --jars, but gives "unresolved dependency" error passed as --packages when read from remote repo (jfrog). But, all other dependencies works like abris, kafka etc.
EDIT- previously asked question with no conclusion: How to use custom jars in spark-submit --packages
We are doing POC on Iceberg and evaluating it first time.
Spark Environment:
Spark Standalone Cluster Setup ( 1 master and 5 workers)
Spark: spark-3.1.2-bin-hadoop3.2
Scala: 2.12.10
Java: 1.8.0_321
Hadoop: 3.2.0
Iceberg 0.13.1
As suggested in Iceberg's official documentation, to add support for Iceberg in Spark shell, we are adding Iceberg dependency while launching the Spark shell as below,
spark-shell --packages org.apache.iceberg:iceberg-spark-runtime-3.2_2.12:0.13.1
After launching the Spark shell with the above command, we are not able to use the Spark shell at all. For all the commands (even non Iceberg) we are getting the same exception as below,
Exception in thread "main" java.lang.NoClassDefFoundError: org/apache/spark/sql/catalyst/plans/logical/BinaryCommand
Below simple command also throwing same exception.
val df : DataFrame = spark.read.json("/spark-3.1.2-bin-hadoop3.2/examples/src/main/resources/people.json")
df.show()
In Spark source code, BinaryCommand class belongs to Spark SQL module, so tried explicitly adding Spark SQL dependency while launching Spark shell as below, but still getting same exception.
spark-shell --packages org.apache.iceberg:iceberg-spark-runtime-3.2_2.12:0.13.1,org.apache.spark:spark-sql_2.12:3.1.2
When we launch spark-shell normally i.e. without Iceberg dependency, then it is working properly.
Any pointer in the right direction for troubleshooting would be really helpful.
Thanks.
We are using the wrong Iceberg version, choose the spark 3.2 iceberg jar but running Spark 3.1. After using the correct dependency version (i.e. 3.1), we are able to launch the Spark shell with Iceberg. Also no need to specify org.apache.spark Spark jars using packages since all of that will be on the classpath anyway.
spark-shell --packages org.apache.iceberg:iceberg-spark-runtime-3.1_2.12:0.13.1
I have a jar file that contains some Scala (and Java) code that I run using the following spark-submit command:
spark-submit
--verbose
--class mycompany.MyClass
--conf spark.driver.extraJavaOptions=-Dconfig.resource=dev-test.conf
--conf "spark.executor.extraJavaOptions=-Dconfig.resource=dev-test.conf -verbose:class"
--conf 'spark.driver.extraJavaOptions=-verbose:class'
--master yarn
--driver-library-path /usr/lib/hadoop-lzo/lib/native/
--jars /home/hadoop/mydir/spark-utils-1.1.0-SNAPSHOT.jar,/usr/lib/phoenix/phoenix-client-hbase-2.4-5.1.2.jar,/usr/lib/hadoop-lzo/lib/hadoop-lzo.jar,/usr/lib/hadoop/lib/commons-compress-1.18.jar,/usr/lib/hadoop/hadoop-aws-3.2.1-amzn-5.jar,/usr/share/aws/aws-java-sdk/aws-java-sdk-bundle-1.12.31.jar
--files /home/hadoop/mydir/dev-test.conf
--queue default /home/hadoop/mydir/spark-utils-1.1.0-SNAPSHOT.jar
<<args to MyClass>>
When I run it, I get an error message - “IAMInstanceCredentialsProvider not found”, which is caused by a version mismatch. It seems IAMInstanceCredentialsProvider was added to hadoop-aws in version 3.3.0 and we want to use 3.2.1. I've gone through our maven dependencies and feel confident that we are not trying to use 3.3.x anywhere.
I've attempted to debug the problem by adding some "verbose" arguments to the command, and I've also added some debug code to MyClass to print out the classpath in effect, following the instructions from here.
When I look at the output, the classpath in effect when we run the spark-submit command includes a lot of jars included with Python, including /usr/local/lib/python3.7/site-packages/pyspark/jars/hadoop-client-api-3.3.1.jar. Thus far, I've been unable to figure out why we are loading jars from /usr/local/lib/python3.7.
Can anybody explain to me where those dependencies are coming from, or suggest a way that I could debug where those dependencies come from? I thought the python might be a result of some environment variable setting, but if so, it doesn't seem to be set at the top level:
set|grep -i python
doesn't return anything.
I've build Spark 2.1 source code successfully.
However, when I run some of the examples (e.g., org.apache.spark.examples.mllib.BinaryClassification), I get the following error.
Exception in thread "main" java.lang.NoClassDefFoundError: scopt/OptionParser
I tried to run those examples using Spark 2.1 pre-built version (examples/jars/spark-examples_2.11-2.1.0.jar), and I got the same error. Spark 1.6 pre-built version works (lib/spark-examples-1.6.2-hadoop2.6.0.jar). There are posts related to this error, but they don't seem to be applicable because Spark examples folder does not have any .sbtfile.
I found the answer. To avoid the error, scopt_x.xx-x.x.x.jar should also be submitted using --jars. When you build Spark examples, in addition to spark-examples_x.xx-x.x.x.jar, scopt_x.xx-x.x.x.jar will be built too (in my case in the same target folder examples/target/scala-2.11/jars).
Once you have the jar file, you can submit it with your applications:
./bin/spark-submit \
--jars examples/target/scala-2.11/jars/scopt_x.xx-x.x.x.jar \
--class org.apache.spark.examples.mllib.BinaryClassification \
--master ...
I have a Spark Job that is using some external libraries to work. When I run the job locally through the main method from IntelliJ the job runs without any issues. However, when I assembly my job into a jarfile (I create an UberJAR using sbt) and I try to run it on EMR, it throws a ClassNotFoundException.
I have checked that the class is indeed inside the jarfile so it should be available for the job to run. I have also tried the spark-submit options spark.driver.extraClassPath, spark.driver.extraLibraryPath, spark.executor.extraClassPath and spark.executor.extraLibraryPath as well as spark.driver.userClassPathFirst and spark.executor.userClassPathFirst. Also, I tried doing in the code sparkContext.addJar("/mnt/jars/myJar"). None of them worked for me.
Also, when running on EMR I can read the log that says that the JAR was added (not sure if it is loaded on the classpath, but it should because other classes are being loaded properly):
15/11/02 04:10:26 INFO SparkContext: Added JAR file:///mnt/my-app-1.0-SNAPSHOT.jar at http://172.31.42.244:44471/jars/my-app-1.0-SNAPSHOT.jar with timestamp 1446437426661
I am running out of ideas about what else to try. I have been researching and I see few tickets on the Spark JIRA board but nothing similar to my issue.
I am running on EMR release-label 4.1.0 (Spark 1.5.0), Java 7, sbt 0.13.7 and Scala 2.10.5.
I think when launching your job on EMR you need to provide the s3 location for your jar dependencies a la the manual e.g. -u s3://sparksupport/libs. These jars will be added to the classpath when running spark.
It turned out to be a problem with SerializationUtils from Apache Commons Lang. There is an open issue where the class will throw a ClassNotFoundException even if the class is in the classpath in a multiple-classloader environment: https://issues.apache.org/jira/browse/LANG-1049
We moved away from the library and our Spark job is working fine now. The issue was not related with Spark finally.