Subject: [prometheus-users] Improving performance of jmx exporter

On Sat, 18 Jan 2020 at 19:27, Julien Pivotto <[EMAIL PROTECTED]> wrote:

The problem is more that JMX is not efficient, and Kafka tends to have high
cardinality due to all the per-partition metrics. Getting kafka to produce
/metrics more directly would be a better approach.
Brian Brazil