mirror of https://github.com/apache/druid.git
2855fb6ff8
* change kafka lookups module to not commit offsets The current behaviour of the Kafka lookup extractor is to not commit offsets by assigning a unique ID to the consumer group and setting auto.offset.reset to earliest. This does the job but also pollutes the Kafka broker with a bunch of "ghost" consumer groups that will never again be used. To fix this, we now set enable.auto.commit to false, which prevents the ghost consumer groups being created in the first place. * update docs to include new enable.auto.commit setting behaviour * update kafka-lookup-extractor documentation Provide some additional detail on functionality and configuration. Hopefully this will make it clearer how the extractor works for developers who aren't so familiar with Kafka. * add comments better explaining the logic of the code * add spelling exceptions for kafka lookup docs |
||
|---|---|---|
| .. | ||
| avro-extensions | ||
| azure-extensions | ||
| datasketches | ||
| druid-aws-rds-extensions | ||
| druid-basic-security | ||
| druid-bloom-filter | ||
| druid-kerberos | ||
| druid-pac4j | ||
| druid-ranger-security | ||
| ec2-extensions | ||
| google-extensions | ||
| hdfs-storage | ||
| histogram | ||
| kafka-extraction-namespace | ||
| kafka-indexing-service | ||
| kinesis-indexing-service | ||
| kubernetes-extensions | ||
| lookups-cached-global | ||
| lookups-cached-single | ||
| mysql-metadata-storage | ||
| orc-extensions | ||
| parquet-extensions | ||
| postgresql-metadata-storage | ||
| protobuf-extensions | ||
| s3-extensions | ||
| simple-client-sslcontext | ||
| stats | ||
| testing-tools | ||