Starting 2019-04-05 it looks like traffic (messages and bytes) to `__consumer_offsets` to logstash1010 kept increasing and stayed at ~15k msg/s. A logstash frontend restart seems to reset the traffic (happened on april 15th, then again on april 16th), although I don't see the traffic increasing now (screenshot below). [link to dashboard](https://grafana.wikimedia.org/d/000000027/kafka?orgId=1&var-datasource=eqiad%20prometheus%2Fops&var-kafka_cluster=logging-eqiad&var-cluster=logstash&var-kafka_broker=All&var-disk_device=sdb&from=1554275664321&to=1555488720321) (note that it'll take a while to load, the dashboard is heavy on prometheus ATM)
I'm filing this for reference / tracking as I don't know if it causes any problems or even what's the root cause. cc @elukey @Ottomata @herron @colewhite
{F28685216}