Druid in clusters can't found other nodes

Hi, im testing the druid (0.12.1) installation in cluster. I have 5 servers with 8gb of ram and 4 cores. Druid is distributed in the next way:

One server with:

  1. coordinator and overlord

  2. zookeeper

  3. historical, midddlemanager, postgresql

  4. broker

  5. and the last one for now host the network file system. In a future i want to use HDFS.

When i do a batch ingestion in the coordinator i see in the logs this thing



so it’s searching the historical in the localhost, when it’s in other server! I don’t know what is wrong…

The task fails one hour later… and there’s no task log…

This is the configuration for all druid units :