Structured Streaming Job stops abruptly and No errors logged
I have a structured streaming job that consumes messages from kafka and does
some stateful associations using flatMapGroupWithState. Every time I submit
the job, it runs fine for around 2hours and then stops abruptly without any
error messages. All I can see in the debug logs is the below message, which
seems to be the starting point that caused the app to shut down.
"INFO SparkContext: Invoking stop() from shutdown hook"
Apart from this message I do not see any exceptions/errors which lead to the
spark context being stopped. Also, I have been monitoring the JMX metrics of
the app using graphite/prometheus exporter. At any point of time, the
memory used never goes beyond the available memory.
Can anyone suggest what could be the issue here.
Also, here is the log4j configuration I am using, does this capture all
types of errors or am I missing something.