Adam Heinz

Reading through https://github.com/nodejs/node/blob/master/lib/_stream_readable.js, specifically push(), readableAddChunk() and addChunk(), I learned that the objects are stored internally in a linked list: > // A linked list is used to store data...

Instantiating QueryStream(text, values, { batchSize: 16 }) seems to have resolved my own memory consumption issues. The batchSize normally defaults to 100; quite a bit higher than the highWaterMark default...

``` $ jar -tf kafka-statsd-metrics2-0.4.1.jar META-INF/ META-INF/MANIFEST.MF com/ com/airbnb/ com/airbnb/kafka/ com/airbnb/kafka/KafkaStatsdMetricsReporter.class com/airbnb/kafka/KafkaStatsdMetricsReporterMBean.class com/airbnb/metrics/ com/airbnb/metrics/Dimension.class com/airbnb/metrics/ExcludeMetricPredicate.class com/airbnb/metrics/MetricNameFormatter.class com/airbnb/metrics/Parser.class com/airbnb/metrics/ParserForNoTag.class com/airbnb/metrics/ParserForTagInMBeanName.class com/airbnb/metrics/StatsDReporter.class ``` suggests ``` kafka.metrics.reporters=com.airbnb.kafka.KafkaStatsdMetricsReporter ``` I'm using the same versions...

Got it partially working. Dropping [java-dogstatsd-client-2.0.11.jar](http://central.maven.org/maven2/com/indeed/java-dogstatsd-client/2.0.11/java-dogstatsd-client-2.0.11.jar) into libs/ and downgrading kafka to 0.8.2.1 (an exact match for [0.4.1/build.gradle#L26](https://github.com/airbnb/kafka-statsd-metrics2/blob/0.4.1/build.gradle#L26)) seems to have most of the metrics showing up, minus the tag-requiring...