Ori Popowski created FLINK-16003:
------------------------------------
Summary: ShardConsumer errors cannot be logged
Key: FLINK-16003
URL: https://issues.apache.org/jira/browse/FLINK-16003
Project: Flink
Issue Type: Bug
Components: Connectors / Kinesis
Affects Versions: 1.6.3
Reporter: Ori Popowski
Some of the errors in the Kinesis connector show on Flink UI, but are not
logged. This causes a serious problem since we cannot see them in our logging
aggregation platform, and we cannot create alerts on them. One of the errors is
the following:
{code:java}
java.lang.RuntimeException: Rate Exceeded for getRecords operation - all 3
retry attempts returned ProvisionedThroughputExceededException.
at
org.apache.flink.streaming.connectors.kinesis.proxy.KinesisProxy.getRecords(KinesisProxy.java:234)
at
org.apache.flink.streaming.connectors.kinesis.internals.ShardConsumer.getRecords(ShardConsumer.java:311)
at
org.apache.flink.streaming.connectors.kinesis.internals.ShardConsumer.run(ShardConsumer.java:219)
at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)
at java.util.concurrent.FutureTask.run(FutureTask.java:266)
at
java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
at
java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
at java.lang.Thread.run(Thread.java:748){code}
The code shows that this exception is not caught and hence is not logged and
can be detected only from the "exceptions" tab in Flink UI. People who use the
connector cannot leverage logging and metrics when such an exception occurs.
It can be useful to catch all the throwables from the ShardConsumer's run()
method and log them.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
