docker-elasticsearch
docker-elasticsearch copied to clipboard
Question on ES nodes discovery in docker-compose
I'm trying to replicate a cluster setup with docker-compose. Please, see docker-compose file below.
version: '3'
services:
elasticsearch-master:
image: quay.io/pires/docker-elasticsearch:6.2.3_1
container_name: elasticsearch-master
privileged: true
environment:
- "HTTP_ENABLE=false"
- "NODE_DATA=false"
- "CLUSTER_NAME=elasticsearch"
- "ES_JAVA_OPTS=-Xms512m -Xmx512m"
networks:
- elastic
elasticsearch:
image: quay.io/pires/docker-elasticsearch:6.2.3_1
container_name: elasticsearch
privileged: true
ports:
- "9200:9200"
- "9300:9300"
environment:
- "NODE_MASTER=false"
- "NODE_DATA=false"
- "CLUSTER_NAME=elasticsearch"
- "ES_JAVA_OPTS=-Xms256m -Xmx256m"
# - "discovery.zen.ping.unicast.hosts=[\"elasticsearch-master\"]"
networks:
- elastic
depends_on:
- elasticsearch-master
elasticsearch-data1:
image: quay.io/pires/docker-elasticsearch:6.2.3_1
container_name: elasticsearch-data1
privileged: true
environment:
- "NODE_MASTER=false"
- "HTTP_ENABLE=false"
- "CLUSTER_NAME=elasticsearch"
- "ES_JAVA_OPTS=-Xms512m -Xmx512m"
- "discovery.zen.ping.unicast.hosts=[\"elasticsearch-master\"]"
depends_on:
- elasticsearch-master
networks:
- elastic
elasticsearch-data2:
image: quay.io/pires/docker-elasticsearch:6.2.3_1
container_name: elasticsearch-data2
privileged: true
environment:
- "NODE_MASTER=false"
- "HTTP_ENABLE=false"
- "CLUSTER_NAME=elasticsearch"
- "ES_JAVA_OPTS=-Xms512m -Xmx512m"
- "discovery.zen.ping.unicast.hosts=[\"elasticsearch-master\"]"
depends_on:
- elasticsearch-master
networks:
- elastic
networks:
elastic:
driver: bridge
But nodes are never able to discover each other with messages like
elasticsearch | [2018-04-20T22:13:56,503][WARN ][o.e.d.z.ZenDiscovery ] [1ee60d4e-9387-47b7-bdcb-2a1ba408826f] not enough master nodes discovered during pinging (found [[]], but needed [1]), pinging again
elasticsearch-data1 | [2018-04-20T22:13:56,890][WARN ][o.e.d.z.ZenDiscovery ] [bf93ab01-324d-4624-9d25-58b6611dcead] not enough master nodes discovered during pinging (found [[]], but needed [1]), pinging again
What am I possibly missing here? Thank you!
I can't help at this point as I don't use docker-compose.