Minor In 9.3 nuxeo.stream.work.log.config point to the LogConfig to use for all the worker pools.
When choosing a Kafka LogConfig it means the same kafka producer/consumer properties for all worker pools.
In the case of pool with slow or fast worker, we may need a different configurations.
For instance when reindexing the entire repository, some work are slow because they are processing lots of documents, this means the max.poll.interval must be high and max.poll.records low. This will not suit very well for fast worker.
Also being able to choose a LogConfig per workpool enable to mix local worker pool (using chronicle queue) and distributed worker pool (using kafka) in the same instance.
