DataStax Ops Center EC2 multi region setup, with 2 nodes in us-east-1 and 2 in us-west-1
Ops center is installed on a separate ec2 instance, and can see the cluster, but I cannot get the agents installed. Looking at one instance, in /var/lib/opscenter-agent/conf/address.yaml, I have:
stomp_interface: "local ip address"
local_interface: "public ip address"
local_address: "public ip address"
use_ssl: 0
agent_rpc_interface: "local ip address"
agent_rpc_broadcast_address: "public ip address"
These are IP addresses, not ec2 hostnames.
On the OpsCenter server side, I have:
[agents]
use_ssl = false
Set in /etc/opscenter/opscenterd.conf.
When I start the agent, I see:
ERROR [StompConnection receiver] 2013-10-14 20:27:49,980 failed connecting to 10.114.79.236:61620:java.net.ConnectException: Connection refused
INFO [StompConnection receiver] 2013-10-14 20:27:49,981 Reconnecting in 60s.
INFO [install-location-finder] 2013-10-14 20:27:56,774 New JMX connection (127.0.0.1:7199)
WARN [clojure-agent-send-off-pool-4] 2013-10-14 20:28:37,083 Tried to send message while not connected: /node-details ["public ip address",{"os-load":0.0}]
OpsCenter can see the details of the cluster (number of nodes, what DC they are in, etc).
In the opscenterd logs, I see:
Node "public ip" has an unknown topology, collecting topology info from another node now
Cycling through all 4 nodes.
I have ports open between the security groups of the agent and server nodes.
Any suggestions would be very welcome.