Cassandra Datastax OpsCenter "NO DATA" on some graphs - cassandra

Question

I'm running on 1 node test cluster Cassandra Datastax OpsCenter 5.2.0, installed from Amazon Datastax AMI version 2.6.3 with Cassandra Community 2.2.0-1.
OpsCenter doesn't report any errors (All agents connected) yet on some graphs I see NO DATA (while I know that there been a lot of requests):
On some there is nothing:
some are working just fine, like OS: Load, Storage Capacity and OS: Disk Utilization.
What could be the reason for this? How to fix it?
EDIT:
Opscenter logs seems to be fine. In agent.log file, I've found the following errors (dozens of times):
ERROR [jmx-metrics-2] 2015-09-21 06:50:30,910 Error getting CF metrics
java.lang.UnsupportedOperationException: nth not supported on this type: PersistentArrayMap
at clojure.lang.RT.nthFrom(RT.java:857)
at clojure.lang.RT.nth(RT.java:807)
at opsagent.rollup$process_metric_map.invoke(rollup.clj:252)
at opsagent.metrics.jmx$cf_metric_helper.invoke(jmx.clj:96)
at opsagent.metrics.jmx$start_pool$fn__15320.invoke(jmx.clj:159)
at clojure.lang.AFn.run(AFn.java:24)
at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:471)
at java.util.concurrent.FutureTask.runAndReset(FutureTask.java:304)
at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.access$301(ScheduledThreadPoolExecutor.java:178)
at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.run(ScheduledThreadPoolExecutor.java:293)
at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1145)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:615)
at java.lang.Thread.run(Thread.java:745)
ERROR [jmx-metrics-4] 2015-09-21 06:50:38,524 Error getting general metrics
java.lang.UnsupportedOperationException: nth not supported on this type: PersistentHashMap
at clojure.lang.RT.nthFrom(RT.java:857)
at clojure.lang.RT.nth(RT.java:807)
at opsagent.rollup$process_metric_map.invoke(rollup.clj:252)
at opsagent.metrics.jmx$generic_metric_helper.invoke(jmx.clj:73)
at opsagent.metrics.jmx$start_pool$fn__15334$fn__15335.invoke(jmx.clj:171)
at opsagent.metrics.jmx$start_pool$fn__15334.invoke(jmx.clj:170)
at clojure.lang.AFn.run(AFn.java:24)
at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:471)
at java.util.concurrent.FutureTask.runAndReset(FutureTask.java:304)
at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.access$301(ScheduledThreadPoolExecutor.java:178)
at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.run(ScheduledThreadPoolExecutor.java:293)
at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1145)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:615)
at java.lang.Thread.run(Thread.java:745)
BTW. In Datastax agent configuration file (address.yaml), I have only stomp_interface parameter set to my node IP.

Answer 1

NO DATA in this case was caused by bug in Datastax agent. I've included related error in my question. The same error is also mentioned here: Opsagent UnsupportedOperationException with PersistentHashMap
After upgrading from Datastax Opscenter 5.2.0 to 5.2.1 and upgrading agent to the same version, problem went away.
Thank you #phact for our help!

Answer 2

I had a similar error before and it was due to a wrongly configured address.yaml.
In my case the issue was that I only wrote a simple IP address in the hosts setting and it was fixed by using an array syntax instead. It seemed that without that, the agent was using a wrong data type, which then issued an UnsupportedOperationException. Try to add your local interface IP (or whatever IP your cassandra node is listening on on your machine) to the hosts setting like so in address.yaml:
hosts: ["<local-node-ip>"]
Maybe also try adding the following settings to address.yaml, just to be sure that the correct parameters are used by the agent:
rpc_address: <local-node-ip>
agent_rpc_interface: <local-node-ip>