[FLINK-24962] The root cause of a failed job was hidden in certain cases

• Type: Improvement
• Status: Closed
• Priority: Major
• Resolution: Cannot Reproduce
• Affects Version/s: 1.14.0
• Component/s: API / Python
• Assignee: Unassigned
• Reporter: Dian Fu
• Created: 19/Nov/21 09:07
• Updated: 07/Jan/22 07:18
• Resolved: 07/Jan/22 07:18

Description

For the following job:

from pyflink.common.typeinfo import Types
from pyflink.datastream import StreamExecutionEnvironment, DataStream
from pyflink.table import StreamTableEnvironment


def state_access_demo():
    env = StreamExecutionEnvironment.get_execution_environment()
    t_env = StreamTableEnvironment.create(stream_execution_environment=env)
    sql = """CREATE TABLE kafka_source (
        id int,
        name string
    ) WITH (
        'connector' = 'kafka',
        'topic' = 'test',
        'properties.bootstrap.servers' = '***',
        'properties.group.id' = '***',
        'scan.startup.mode' = 'latest-offset',
        'format' = 'json',
        'json.fail-on-missing-field' = 'false',
        'json.ignore-parse-errors' = 'true'
    )"""
    t_env.execute_sql(sql)
    table = t_env.from_path("kafka_source")
    ds: DataStream = t_env.to_append_stream(table, type_info=Types.ROW([Types.INT(), Types.STRING()]))
    ds.map(lambda a: print(a))
    env.execute('state_access_demo')

if __name__ == '__main__':
    state_access_demo()

It failed with the following exception which doesn't contain any useful information (the root cause is that it should return a value in the map function):

Caused by: java.lang.RuntimeException: Failed to create stage bundle factory! INFO:root:Initializing Python harness: D:\soft\anaconda\lib\site-packages\pyflink\fn_execution\beam\beam_boot.py --id=2-1 --provision_endpoint=localhost:57201
INFO:root:Starting up Python harness in loopback mode.

    at org.apache.flink.streaming.api.runners.python.beam.BeamPythonFunctionRunner.createStageBundleFactory(BeamPythonFunctionRunner.java:566)
    at org.apache.flink.streaming.api.runners.python.beam.BeamPythonFunctionRunner.open(BeamPythonFunctionRunner.java:255)
    at org.apache.flink.streaming.api.operators.python.AbstractPythonFunctionOperator.open(AbstractPythonFunctionOperator.java:131)
    at org.apache.flink.streaming.api.operators.python.AbstractOneInputPythonFunctionOperator.open(AbstractOneInputPythonFunctionOperator.java:116)
    at org.apache.flink.streaming.api.operators.python.PythonProcessOperator.open(PythonProcessOperator.java:59)
    at org.apache.flink.streaming.runtime.tasks.RegularOperatorChain.initializeStateAndOpenOperators(RegularOperatorChain.java:110)
    at org.apache.flink.streaming.runtime.tasks.StreamTask.restoreGates(StreamTask.java:711)
    at org.apache.flink.streaming.runtime.tasks.StreamTaskActionExecutor$1.call(StreamTaskActionExecutor.java:55)
    at org.apache.flink.streaming.runtime.tasks.StreamTask.restoreInternal(StreamTask.java:687)
    at org.apache.flink.streaming.runtime.tasks.StreamTask.restore(StreamTask.java:654)
    at org.apache.flink.runtime.taskmanager.Task.runWithSystemExitMonitoring(Task.java:958)
    at org.apache.flink.runtime.taskmanager.Task.restoreAndInvoke(Task.java:927)
    at org.apache.flink.runtime.taskmanager.Task.doRun(Task.java:766)
    at org.apache.flink.runtime.taskmanager.Task.run(Task.java:575)
    at java.lang.Thread.run(Thread.java:748)
Caused by: org.apache.beam.vendor.guava.v26_0_jre.com.google.common.util.concurrent.UncheckedExecutionException: java.lang.IllegalStateException: Process died with exit code 0
    at org.apache.beam.vendor.guava.v26_0_jre.com.google.common.cache.LocalCache$Segment.get(LocalCache.java:2050)
    at org.apache.beam.vendor.guava.v26_0_jre.com.google.common.cache.LocalCache.get(LocalCache.java:3952)
    at org.apache.beam.vendor.guava.v26_0_jre.com.google.common.cache.LocalCache.getOrLoad(LocalCache.java:3974)
    at org.apache.beam.vendor.guava.v26_0_jre.com.google.common.cache.LocalCache$LocalLoadingCache.get(LocalCache.java:4958)
    at org.apache.beam.vendor.guava.v26_0_jre.com.google.common.cache.LocalCache$LocalLoadingCache.getUnchecked(LocalCache.java:4964)
    at org.apache.beam.runners.fnexecution.control.DefaultJobBundleFactory$SimpleStageBundleFactory.<init>(DefaultJobBundleFactory.java:451)
    at org.apache.beam.runners.fnexecution.control.DefaultJobBundleFactory$SimpleStageBundleFactory.<init>(DefaultJobBundleFactory.java:436)
    at org.apache.beam.runners.fnexecution.control.DefaultJobBundleFactory.forStage(DefaultJobBundleFactory.java:303)
    at org.apache.flink.streaming.api.runners.python.beam.BeamPythonFunctionRunner.createStageBundleFactory(BeamPythonFunctionRunner.java:564)
    ... 14 more
Caused by: java.lang.IllegalStateException: Process died with exit code 0
    at org.apache.beam.runners.fnexecution.environment.ProcessManager$RunningProcess.isAliveOrThrow(ProcessManager.java:75)
    at org.apache.beam.runners.fnexecution.environment.ProcessEnvironmentFactory.createEnvironment(ProcessEnvironmentFactory.java:112)
    at org.apache.beam.runners.fnexecution.control.DefaultJobBundleFactory$1.load(DefaultJobBundleFactory.java:252)
    at org.apache.beam.runners.fnexecution.control.DefaultJobBundleFactory$1.load(DefaultJobBundleFactory.java:231)
    at org.apache.beam.vendor.guava.v26_0_jre.com.google.common.cache.LocalCache$LoadingValueReference.loadFuture(LocalCache.java:3528)
    at org.apache.beam.vendor.guava.v26_0_jre.com.google.common.cache.LocalCache$Segment.loadSync(LocalCache.java:2277)
    at org.apache.beam.vendor.guava.v26_0_jre.com.google.common.cache.LocalCache$Segment.lockedGetOrLoad(LocalCache.java:2154)
    at org.apache.beam.vendor.guava.v26_0_jre.com.google.common.cache.LocalCache$Segment.get(LocalCache.java:2044)

It's very difficult for users to figure out why the job failed and we should improve this.