Why Cassandra nodes do not join cluster?

Question

I am trying to create two nodes of Cassandra, both nodes will start, but when I checknodetool status I get:

On first node:

$ nodetool status
Datacenter: eu-central
======================
Status=Up/Down
|/ State=Normal/Leaving/Joining/Moving
--  Address     Load       Tokens       Owns (effective)  Host ID                               Rack
UN  10.0.30.83  217.7 KiB  256          100.0%            38068f28-5116-4bd9-89a4-5787a112dd08  1b

On other node: $ nodetool status

Datacenter: eu-central
======================
Status=Up/Down
|/ State=Normal/Leaving/Joining/Moving
--  Address     Load       Tokens       Owns (effective)  Host ID                               Rack
UN  10.0.5.247  142.96 KiB  256          100.0%            0667775c-1365-4696-baa7-ce15fea186b3  1a

both nodes has seeds in cassandra.yaml as:

- seeds: "10.0.5.247,10.0.30.83"

both nodes can ping each other, and as for port 7000:

first node:

netstat -anp | grep 7000
(No info could be read for "-p": geteuid()=1000 but you should be root.) tcp        0      0 127.0.0.1:7000          0.0.0.0:*

        LISTEN      -

second node:

$ netstat -anp | grep 7000
(No info could be read for "-p": geteuid()=1000 but you should be root.)
tcp        0      0 127.0.0.1:7000          0.0.0.0:*               LISTEN      -

So, why the do not connect to each other?

I agree with Chris. Check the system.log for errors about handshaking with the other node. — Aaron
It was fixed by setting listen_address to private ip of node — simo
Also, node was started by simple snitch, then when I tried EC2 snitch, I didn't delete data directory, which also was an issue — simo

nevsv nevsv · Accepted Answer · 2017-06-01T14:47:40

Verify that they both have same cluster name.
Check rpc_adress (0.0.0.0 for example) and listen_address(should be the one you've listed in the seeds list) on both nodes.

Why Cassandra nodes do not join cluster?

2 Answers