Patrick McGloin created SPARK-25466:
---------------------------------------
Summary: Documentation does not specify how to set fafka consumer
cache capacity for SS
Key: SPARK-25466
URL: https://issues.apache.org/jira/browse/SPARK-25466
Project: Spark
Issue Type: Bug
Components: Structured Streaming
Affects Versions: 2.3.0
Reporter: Patrick McGloin
When hitting this warning with SS:
19-09-2018 12:05:27 WARN CachedKafkaConsumer:66 - KafkaConsumer cache hitting
max capacity of 64, removing consumer for
CacheKey(spark-kafka-source-e06c9676-32c6-49c4-80a9-2d0ac4590609--694285871-executor,MyKafkaTopic-30)
If you Google you get to this page:
https://spark.apache.org/docs/latest/streaming-kafka-0-10-integration.html
Which is for Spark Streaming and says to use this config item to adjust the
capacity: "spark.streaming.kafka.consumer.cache.maxCapacity".
This is a bit confusing as SS uses a different config item:
"spark.sql.kafkaConsumerCache.capacity"
Perhaps the SS Kafka documentation should talk about the consumer cache
capacity? Perhaps here?
https://spark.apache.org/docs/2.2.0/structured-streaming-kafka-integration.html
Or perhaps the warning message should reference the config item. E.g
19-09-2018 12:05:27 WARN CachedKafkaConsumer:66 - KafkaConsumer cache hitting
max capacity of 64, removing consumer for
CacheKey(spark-kafka-source-e06c9676-32c6-49c4-80a9-2d0ac4590609--694285871-executor,MyKafkaTopic-30).
*The cache size can be adjusted with the setting
"spark.sql.kafkaConsumerCache.capacity".*
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
