5

I'm setting up druid first time and ran into following issues while trying to start druid using docker-compose,

2020-04-10T14:40:01,837 ERROR [qtp1667348377-84] org.apache.druid.server.router.QueryHostFinder - Catastrophic failure! No servers found at all! Failing request!: {class=org.apache.druid.server.router.QueryHostFinder} 2020-04-10T14:40:01,837 WARN [qtp1667348377-84] org.eclipse.jetty.server.HttpChannel - /druid/v2/sql org.apache.druid.java.util.common.ISE: No default server found! at org.apache.druid.server.router.QueryHostFinder.pickDefaultServer(QueryHostFinder.java:119) ~[druid-server-0.17.1.jar:0.17.1]

here is the cmd i'm using

vm:~$ sudo docker-compose -f distribution/docker/docker-compose.yml up

i've cloned the repo from https://github.com/apache/druid.

However if download the druid distribution apache-druid-0.17.1 and using ./bin/start-micro-quickstart I'm able to see all services started successfully and running fine. able to access web-console and load the data into segments.

But when I try to start druid in cluster mode or using docker-compose getting 404 errors in console and in logs connection refused and Out Of Memory errors. I've increased -XX:MaxDirectMemorySize=6g in druid.sh but no luck.

Please help me in resolving these errors while starting druid using docker-compose.

marc_s
  • 732,580
  • 175
  • 1,330
  • 1,459
Mnd
  • 51
  • 1
  • 2
  • Similar issue here. The broker seems to exit with code 137, which maps to some process running `kill -9` on it. http://tldp.org/LDP/abs/html/exitcodes.html – TheLonelyGhost Apr 23 '20 at 16:43

2 Answers2

4

Do a Docker PS to check if router/broker is running.

If not, increase the memory allocated to your docker engine. Druid need at least 4GB memory.

Kai_90
  • 65
  • 6
  • 1
    docker ps shows all the following services are up middlemanager,broker,router,historical,coordinator,zookeeper,postgres I've increased the DirectMemorySize to 6GB and -Xmx to 7G – Mnd Apr 17 '20 at 11:56
  • 3
    this solved my issue, unlike @Mnd the services that were up for me were: middlemanager,router,coordinator,zookeeper,postgres. I increased the resources from 2gb to 8gb. and for reference my cpus is at 4, swag 1gb, disk image size 60 gb. if you want to update your memory via docker desktop go to preferences > resources. – Snackdex May 21 '20 at 20:42
0

I had the same issue. docker restart coordinator fixed it.

Daniel Darabos
  • 26,991
  • 10
  • 102
  • 114