Druid 0.11.0 historical GC Zk client timeout

The issue happens when the historical jvm GC beginning

Below is the jvm/gc/cpu , the value 47992773866 meaning 47s? Out zookeeper client timeout is 30s, so the node is kicked out。

superuser@druid-historical-001:/app/druid/log/druid$ grep ‘jvm/gc/cpu’ historical-2018-06-08*

historical-2018-06-08.5.log:2018-06-08 15:12:42,341 INFO [MonitorScheduler-0] com.metamx.emitter.core.LoggingEmitter - Event [{“feed”:“metrics”,“timestamp”:“2018-06-08T07:12:42.341Z”,“service”:“druid/historical”,“host”:“druid-historical-001:8083”,“version”:“0.11.0”,“metric”:“jvm/gc/cpu”,“value”:11663165650,“gcGen”:[“young”],“gcName”:[“g1”]}]

historical-2018-06-08.5.log:2018-06-08 15:12:42,341 INFO [MonitorScheduler-0] com.metamx.emitter.core.LoggingEmitter - Event [{“feed”:“metrics”,“timestamp”:“2018-06-08T07:12:42.341Z”,“service”:“druid/historical”,“host”:“druid-historical-001:8083”,“version”:“0.11.0”,“metric”:“jvm/gc/cpu”,“value”:17773099492,“gcGen”:[“old”],“gcName”:[“g1”]}]

historical-2018-06-08.5.log:2018-06-08 15:14:00,436 INFO [MonitorScheduler-0] com.metamx.emitter.core.LoggingEmitter - Event [{“feed”:“metrics”,“timestamp”:“2018-06-08T07:14:00.436Z”,“service”:“druid/historical”,“host”:“druid-historical-001:8083”,“version”:“0.11.0”,“metric”:“jvm/gc/cpu”,“value”:12107818227,“gcGen”:[“young”],“gcName”:[“g1”]}]

historical-2018-06-08.5.log:2018-06-08 15:14:00,437 INFO [MonitorScheduler-0] com.metamx.emitter.core.LoggingEmitter - Event [{“feed”:“metrics”,“timestamp”:“2018-06-08T07:14:00.437Z”,“service”:“druid/historical”,“host”:“druid-historical-001:8083”,“version”:“0.11.0”,“metric”:“jvm/gc/cpu”,“value”:47992773866,“gcGen”:[“old”],“gcName”:[“g1”]}]

historical-2018-06-08.5.log:2018-06-08 15:15:08,505 INFO [MonitorScheduler-0] com.metamx.emitter.core.LoggingEmitter - Event [{“feed”:“metrics”,“timestamp”:“2018-06-08T07:15:08.505Z”,“service”:“druid/historical”,“host”:“druid-historical-001:8083”,“version”:“0.11.0”,“metric”:“jvm/gc/cpu”,“value”:13228874523,“gcGen”:[“young”],“gcName”:[“g1”]}]

historical-2018-06-08.5.log:2018-06-08 15:15:08,506 INFO [MonitorScheduler-0] com.metamx.emitter.core.LoggingEmitter - Event [{“feed”:“metrics”,“timestamp”:“2018-06-08T07:15:08.506Z”,“service”:“druid/historical”,“host”:“druid-historical-001:8083”,“version”:“0.11.0”,“metric”:“jvm/gc/cpu”,“value”:48096330431,“gcGen”:[“old”],“gcName”:[“g1”]}]

historical-2018-06-08.7.log:2018-06-08 15:16:08,505 INFO [MonitorScheduler-0] com.metamx.emitter.core.LoggingEmitter - Event [{“feed”:“metrics”,“timestamp”:“2018-06-08T07:16:08.505Z”,“service”:“druid/historical”,“host”:“druid-historical-001:8083”,“version”:“0.11.0”,“metric”:“jvm/gc/cpu”,“value”:7355174245,“gcGen”:[“young”],“gcName”:[“g1”]}]

historical-2018-06-08.7.log:2018-06-08 15:16:08,509 INFO [MonitorScheduler-0] com.metamx.emitter.core.LoggingEmitter - Event [{“feed”:“metrics”,“timestamp”:“2018-06-08T07:16:08.509Z”,“service”:“druid/historical”,“host”:“druid-historical-001:8083”,“version”:“0.11.0”,“metric”:“jvm/gc/cpu”,“value”:48424754308,“gcGen”:[“old”],“gcName”:[“g1”]}]

jvm config:

superus+ 27883 1 9 Jun07 ? 01:54:58 java -server -Xms8g -Xmx8g -XX:+UseG1GC -XX:+PrintGCDetails -XX:+PrintGCTimeStamps -XX:+UseGCLogFileRotation -XX:+HeapDumpOnOutOfMemoryError -XX:MaxDirectMemorySize=19327352832 -Duser.timezone=UTC -Dfile.encoding=UTF-8 -Djava.io.tmpdir=var/tmp -Djava.util.logging.manager=org.apache.logging.log4j.jul.LogManager -cp conf/druid/historical:conf/druid/_common:lib/* io.druid.cli.Main server historical