Hi, I would like to open an issue about creating a cluster with multiple machine (instance) that exist in same subnet.
I know this question is kind of duplicated from issue #94 , but after I read all the response, I still get very confused.
Assume that I enabled the swarm mode and let other machine join the swarm as worker, and also, I create the overlay network by sudo docker network create --driver overlay --scope swarm cassandra-net
Indeed, I tried many different yaml file from internet, such as this one. However, when I tried to compose it with docker stack deploy, all the services failed to create node
ID NAME MODE REPLICAS IMAGE PORTS fz0p7gvutwip cassandra_cassandra-1 replicated 0/1 cassandra:latest *:30010->7000/tcp,*:30011->7199/tcp,*:30012->7001/tcp,*:30013->9160/tcp,*:30014->9042/tcp mv49l208lq57 cassandra_cassandra-2 replicated 0/1 cassandra:latest *:30000->7000/tcp,*:30001->7199/tcp,*:30002->7001/tcp,*:30003->9160/tcp,*:30004->9042/tcp xveitbg29boc cassandra_cassandra-3 replicated 0/1 cassandra:latest *:30005->7000/tcp,*:30006->7199/tcp,*:30007->7001/tcp,*:30008->9160/tcp,*:30009->9042/tcp
Then, I tried manually create the service by
sudo docker service create --replicas 3 --name "tesetCluster" --network cassandra-net cassandra
the services can successfully create 3 cassandra container and use the network cassandra-net
But, after I check the nodetool status sudo docker exec -it e4f77b71a6fa nodetool status
I know that I didn't modify the seeds and broadcast_seeds for each cassandra container, but I don't know how to setup the seed for all containers that I created by using service and make those containers point to (see) each other to make a cluster.
And could I ask how to create the correct yaml file for stack deploy to do that.
Thank you very much.
PS. all ports for cassandra opened in each instance (since they use same security group)
yosifkit
commented
6 years ago
Hi, I would like to open an issue about creating a cluster with multiple machine (instance) that exist in same subnet.
I know this question is kind of duplicated from issue #94 , but after I read all the response, I still get very confused.
Assume that I enabled the swarm mode and let other machine join the swarm as worker, and also, I create the overlay network by
sudo docker network create --driver overlay --scope swarm cassandra-netIndeed, I tried many different yaml file from internet, such as this one. However, when I tried to compose it with
docker stack deploy, all the services failed to create nodeID NAME MODE REPLICAS IMAGE PORTSfz0p7gvutwip cassandra_cassandra-1 replicated 0/1 cassandra:latest *:30010->7000/tcp,*:30011->7199/tcp,*:30012->7001/tcp,*:30013->9160/tcp,*:30014->9042/tcpmv49l208lq57 cassandra_cassandra-2 replicated 0/1 cassandra:latest *:30000->7000/tcp,*:30001->7199/tcp,*:30002->7001/tcp,*:30003->9160/tcp,*:30004->9042/tcpxveitbg29boc cassandra_cassandra-3 replicated 0/1 cassandra:latest *:30005->7000/tcp,*:30006->7199/tcp,*:30007->7001/tcp,*:30008->9160/tcp,*:30009->9042/tcpThen, I tried manually create the service by
sudo docker service create --replicas 3 --name "tesetCluster" --network cassandra-net cassandrathe services can successfully create 3 cassandra container and use the network cassandra-net But, after I check the nodetool statussudo docker exec -it e4f77b71a6fa nodetool statusI only got one node for that container
Datacenter: datacenter1Status=Up/Down|/ State=Normal/Leaving/Joining/Moving-- Address Load Tokens Owns (effective) Host ID RackUN 10.0.0.25 103.69 KiB 256 100.0% 9732c109-42ad-4d39-8d4c-d631bd7e8237 rack1I know that I didn't modify the seeds and broadcast_seeds for each cassandra container, but I don't know how to setup the seed for all containers that I created by using service and make those containers point to (see) each other to make a cluster.
And could I ask how to create the correct yaml file for stack deploy to do that.
Thank you very much. PS. all ports for cassandra opened in each instance (since they use same security group)