I have a loop in a pyspark (Spark3 cluster) task like this :
def myfunc(rows):
#Some dynamodb Table initiation stuff (nothing fancy)
with table.batch_writer() as batch:
for row in rows():
try:
batch.put_item(..)
except ClientError as e:
if e.response['Error']['Code'] == "ProvisionedThroughputExceededException":
#handle the issue here
And here is the call to this function from spark
df.foreachPartition(lambda x : myfunc(x))
This code actually works fine. Sometime I receive the exception ProvisionedThroughputExceededException and it's handled. However something super weird is that, if the task handling the bunch of rows seems to encounter the exception it will end as a failing task eventhough the excpetion has been handled, as if spark task check some kind of historical exception to see if something bad happend during the processing:
Here the output from the task :
Getting An error occurred (ProvisionedThroughputExceededException) when calling the BatchWriteItem operation ... ==> handling
Getting An error occurred (ProvisionedThroughputExceededException) when calling the BatchWriteItem operation ... ==> handling
Getting An error occurred (ProvisionedThroughputExceededException) when calling the BatchWriteItem operation ... ==> handling
2022-03-30 08:40:33,029 ERROR Executor: Exception in task 0.0 in stage 2.0 (TID 9)
and then it prints out the stack trace as follows
org.apache.spark.api.python.PythonException: Traceback (most recent call last):
File "/srv/ssd2/yarn/nm/usercache/svc_df_omni/appcache/application_1648119616278_365920/container_e298_1648119616278_365920_01_000004/pyspark.zip/pyspark/worker.py", line 605, in main
process()
File "/srv/ssd2/yarn/nm/usercache/svc_df_omni/appcache/application_1648119616278_365920/container_e298_1648119616278_365920_01_000004/pyspark.zip/pyspark/worker.py", line 595, in process
out_iter = func(split_index, iterator)
File "/srv/ssd2/yarn/nm/usercache/svc_df_omni/appcache/application_1648119616278_365920/container_e298_1648119616278_365920_01_000001/pyspark.zip/pyspark/rdd.py", line 2596, in pipeline_func
File "/srv/ssd2/yarn/nm/usercache/svc_df_omni/appcache/application_1648119616278_365920/container_e298_1648119616278_365920_01_000001/pyspark.zip/pyspark/rdd.py", line 2596, in pipeline_func
File "/srv/ssd2/yarn/nm/usercache/svc_df_omni/appcache/application_1648119616278_365920/container_e298_1648119616278_365920_01_000001/pyspark.zip/pyspark/rdd.py", line 2596, in pipeline_func
File "/srv/ssd2/yarn/nm/usercache/svc_df_omni/appcache/application_1648119616278_365920/container_e298_1648119616278_365920_01_000001/pyspark.zip/pyspark/rdd.py", line 425, in func
File "/srv/ssd2/yarn/nm/usercache/svc_df_omni/appcache/application_1648119616278_365920/container_e298_1648119616278_365920_01_000001/pyspark.zip/pyspark/rdd.py", line 874, in func
File "6.YL_flow_2-ecf3d86.py", line 136, in <lambda>
File "6.YL_flow_2-ecf3d86.py", line 98, in greedy_dyn_send
File "/srv/ssd2/yarn/nm/usercache/svc_df_omni/appcache/application_1648119616278_365920/container_e298_1648119616278_365920_01_000004/env/lib/python3.7/site-packages/boto3/dynamodb/table.py", line 156, in __exit__
self._flush()
File "/srv/ssd2/yarn/nm/usercache/svc_df_omni/appcache/application_1648119616278_365920/container_e298_1648119616278_365920_01_000004/env/lib/python3.7/site-packages/boto3/dynamodb/table.py", line 137, in _flush
RequestItems={self._table_name: items_to_send})
File "/srv/ssd2/yarn/nm/usercache/svc_df_omni/appcache/application_1648119616278_365920/container_e298_1648119616278_365920_01_000004/env/lib/python3.7/site-packages/botocore/client.py", line 388, in _api_call
return self._make_api_call(operation_name, kwargs)
File "/srv/ssd2/yarn/nm/usercache/svc_df_omni/appcache/application_1648119616278_365920/container_e298_1648119616278_365920_01_000004/env/lib/python3.7/site-packages/botocore/client.py", line 708, in _make_api_call
raise error_class(parsed_response, operation_name)
botocore.errorfactory.ProvisionedThroughputExceededException: An error occurred (ProvisionedThroughputExceededException) when calling the BatchWriteItem operation (reached max retries: 1): The level of configured provisioned throughput for the table was exceeded. Consider increasing your provisioning level with the UpdateTable API.
at org.apache.spark.api.python.BasePythonRunner$ReaderIterator.handlePythonException(PythonRunner.scala:503)
at org.apache.spark.api.python.PythonRunner$$anon$3.read(PythonRunner.scala:638)
at org.apache.spark.api.python.PythonRunner$$anon$3.read(PythonRunner.scala:621)
at org.apache.spark.api.python.BasePythonRunner$ReaderIterator.hasNext(PythonRunner.scala:456)
at org.apache.spark.InterruptibleIterator.hasNext(InterruptibleIterator.scala:37)
at scala.collection.Iterator.foreach(Iterator.scala:941)
at scala.collection.Iterator.foreach$(Iterator.scala:941)
at org.apache.spark.InterruptibleIterator.foreach(InterruptibleIterator.scala:28)
at scala.collection.generic.Growable.$plus$plus$eq(Growable.scala:62)
at scala.collection.generic.Growable.$plus$plus$eq$(Growable.scala:53)
at scala.collection.mutable.ArrayBuffer.$plus$plus$eq(ArrayBuffer.scala:105)
at scala.collection.mutable.ArrayBuffer.$plus$plus$eq(ArrayBuffer.scala:49)
at scala.collection.TraversableOnce.to(TraversableOnce.scala:315)
at scala.collection.TraversableOnce.to$(TraversableOnce.scala:313)
at org.apache.spark.InterruptibleIterator.to(InterruptibleIterator.scala:28)
at scala.collection.TraversableOnce.toBuffer(TraversableOnce.scala:307)
at scala.collection.TraversableOnce.toBuffer$(TraversableOnce.scala:307)
at org.apache.spark.InterruptibleIterator.toBuffer(InterruptibleIterator.scala:28)
at scala.collection.TraversableOnce.toArray(TraversableOnce.scala:294)
at scala.collection.TraversableOnce.toArray$(TraversableOnce.scala:288)
at org.apache.spark.InterruptibleIterator.toArray(InterruptibleIterator.scala:28)
at org.apache.spark.rdd.RDD.$anonfun$collect$2(RDD.scala:1004)
at org.apache.spark.SparkContext.$anonfun$runJob$5(SparkContext.scala:2154)
at org.apache.spark.scheduler.ResultTask.runTask(ResultTask.scala:90)
at org.apache.spark.scheduler.Task.run(Task.scala:127)
at org.apache.spark.executor.Executor$TaskRunner.$anonfun$run$3(Executor.scala:462)
at org.apache.spark.util.Utils$.tryWithSafeFinally(Utils.scala:1377)
at org.apache.spark.executor.Executor$TaskRunner.run(Executor.scala:465)
at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
at java.lang.Thread.run(Thread.java:748)
2022-03-30 08:40:33,089 INFO YarnCoarseGrainedExecutorBackend: Got assigned task 73
So I was wondering how spark handle "finishing" a task. Will it say that the task is failed if we encounter an exception and handled it ? Should we clean something whenever we handle an exception ?